Fine-tuning LLMs often requires extensive resources, time, and memory, challenges that can hinder rapid experimentation and deployment. Unsloth AI revolutionizes this process by enabling fast, efficient fine-tuning state-of-the-art models like Qwen3-14B with minimal GPU memory, leveraging advanced techniques such as 4-bit quantization and LoRA. In this tutorial, we walk through a practical implementation on Google Colab to fine-tune Qwen3-14B using a combination of reasoning and instruction-following datasets, combining Unsloth’s FastLanguageModel utilities with trl.SFTTrainer users can achieve powerful fine-tuning performance with just consumer-grade hardware. %%capture import os if "COLAB_" not in "".join): !pip install unsloth else: !pip install --no-deps bitsandbytes accelerate xformers==0.0.29.post3 peft trl==0.15.2 triton cut_cross_entropy unsloth_zoo !pip install sentencepiece protobuf "datasets>=3.4.1" huggingface_hub hf_transfer !pip install --no-deps unsloth We install all the essential libraries required for fine-tuning the Qwen3 model using Unsloth AI. It conditionally installs dependencies based on the environment, using a lightweight approach on Colab to ensure compatibility and reduce overhead. Key components like bitsandbytes, trl, xformers, and unsloth_zoo are included to enable 4-bit quantized training and LoRA-based optimization. from unsloth import FastLanguageModel import torch model, tokenizer = FastLanguageModel.from_pretrainedWe load the Qwen3-14B model using FastLanguageModel from the Unsloth library, which is optimized for efficient fine-tuning. It initializes the model with a context length of 2048 tokens and loads it in 4-bit precision, significantly reducing memory usage. Full fine-tuning is disabled, making it suitable for lightweight parameter-efficient techniques like LoRA. model = FastLanguageModel.get_peft_modelWe apply LoRAto the Qwen3 model using FastLanguageModel.get_peft_model. It injects trainable adapters into specific transformer layerswith a rank of 32, enabling efficient fine-tuning while keeping most model weights frozen. Using “unsloth” gradient checkpointing further optimizes memory usage, making it suitable for training large models on limited hardware. from datasets import load_dataset reasoning_dataset = load_datasetnon_reasoning_dataset = load_datasetWe load two pre-curated datasets from the Hugging Face Hub using the library. The reasoning_dataset contains chain-of-thoughtproblems from Unsloth’s OpenMathReasoning-mini, designed to enhance logical reasoning in the model. The non_reasoning_dataset pulls general instruction-following data from mlabonne’s FineTome-100k, which helps the model learn broader conversational and task-oriented skills. Together, these datasets support a well-rounded fine-tuning objective. def generate_conversation: problems = examplessolutions = examplesconversations =for problem, solution in zip: conversations.appendreturn {"conversations": conversations} This function, generate_conversation, transforms raw question–answer pairs from the reasoning dataset into a chat-style format suitable for fine-tuning. For each problem and its corresponding generated solution, a conversation is conducted in which the user asks a question and the assistant provides the answer. The output is a list of dictionaries following the structure expected by chat-based language models, preparing the data for tokenization with a chat template. reasoning_conversations = tokenizer.apply_chat_templatefrom unsloth.chat_templates import standardize_sharegpt dataset = standardize_sharegptnon_reasoning_conversations = tokenizer.apply_chat_templateimport pandas as pd chat_percentage = 0.75 non_reasoning_subset = pd.Series.sample*), random_state=2407, ) data = pd.concat, pd.Series]) data.name = "text" We prepare the fine-tuning dataset by converting the reasoning and instruction datasets into a consistent chat format and then combining them. It first applies the tokenizer’s apply_chat_template to convert structured conversations into tokenizable strings. The standardize_sharegpt function normalizes the instruction dataset into a compatible structure. Then, a 75-25 mix is created by sampling 25% of the non-reasoningconversations and combining them with the reasoning data. This blend ensures the model is exposed to logical reasoning and general instruction-following tasks, improving its versatility during training. The final combined data is stored as a single-column Pandas Series named “text”. from datasets import Dataset combined_dataset = Dataset.from_pandas) combined_dataset = combined_dataset.shufflefrom trl import SFTTrainer, SFTConfig trainer = SFTTrainer) We take the preprocessed conversations, wrap them into a Hugging Face Dataset, and shuffle the dataset with a fixed seed for reproducibility. Then, the fine-tuning trainer is initialized using trl’s SFTTrainer and SFTConfig. The trainer is set up to use the combined datasetand defines training hyperparameters like batch size, gradient accumulation, number of warmup and training steps, learning rate, optimizer parameters, and a linear learning rate scheduler. This configuration is geared towards efficient fine-tuning while maintaining reproducibility and logging minimal details. trainer.traintrainer.trainstarts the fine-tuning process for the Qwen3-14B model using the SFTTrainer. It trains the model on the prepared mixed dataset of reasoning and instruction-following conversations, optimizing only the LoRA-adapted parameters thanks to the underlying Unsloth setup. Training will proceed according to the configuration specified earlier, and progress will be printed every logging step. This final command launches the actual model adaptation based on your custom data. model.save_pretrainedtokenizer.save_pretrainedWe save the fine-tuned model and tokenizer locally to the “qwen3-finetuned-colab” directory. By calling save_pretrained, the adapted weights and tokenizer configuration can be reloaded later for inference or further training, locally or for uploading to the Hugging Face Hub. In conclusion, with the help of Unsloth AI, fine-tuning massive LLMs like Qwen3-14B becomes feasible, using limited resources, and is highly efficient and accessible. This tutorial demonstrated how to load a 4-bit quantized version of the model, apply structured chat templates, mix multiple datasets for better generalization, and train using TRL’s SFTTrainer. Whether you’re building custom assistants or specialized domain models, Unsloth’s tools dramatically reduce the barrier to fine-tuning at scale. As open-source fine-tuning ecosystems evolve, Unsloth continues to lead the way in making LLM training faster, cheaper, and more practical for everyone. Check out the COLAB NOTEBOOK. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 95k+ ML SubReddit and Subscribe to our Newsletter. Asif RazzaqWebsite |  + postsBioAsif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over 2 million monthly views, illustrating its popularity among audiences.Asif Razzaqhttps://www.marktechpost.com/author/6flvq/Chain-of-Thought May Not Be a Window into AI’s Reasoning: Anthropic’s New Study Reveals Hidden GapsAsif Razzaqhttps://www.marktechpost.com/author/6flvq/How to Build a Powerful and Intelligent Question-Answering System by Using Tavily Search API, Chroma, Google Gemini LLMs, and the LangChain FrameworkAsif Razzaqhttps://www.marktechpost.com/author/6flvq/AWS Open-Sources Strands Agents SDK to Simplify AI Agent DevelopmentAsif Razzaqhttps://www.marktechpost.com/author/6flvq/Windsurf Launches SWE-1: A Frontier AI Model Family for End-to-End Software Engineering 🚨 Build GenAI you can trust. ⭐️ Parlant is your open-source engine for controlled, compliant, and purposeful AI conversations — Star Parlant on GitHub! #stepbystep #coding #guide #efficiently #finetune

Since President Donald Trump excitedly announced that he would be accepting a million plane from the Qatari government to serve as the next Air Force One, even members of his own party have expressed alarm. There’s the price tag of refurbishing the plane with top-secret systems—upward of billion, according to some estimates. Then there are the conflicts of interest from accepting such a large present from a foreign nation—what some say would be the most valuable gift ever given to the U.S. But it would also mark a striking departure from tradition. While they’re often variants of commercial planes, presidential planes have almost always been U.S. military aircraft, flown and maintained by the Air Force. The first White Houses in the sky I’m an aviation historian who once worked in the United States Air Force’s history program for three years, so I’m well acquainted with the history of presidential aircraft. Franklin D. Roosevelt became the first president to fly while in office. In January 1943, he boarded the Navy-owned, civilian-operated Boeing Dixie Clipper—a seaplane—for a trip to Casablanca to meet with Allied leaders. President Franklin D. Roosevelt made the first presidential flight on a Dixie Clipper, a seaplane built by Boeing.The security measures needed to safely transport the president—especially during wartime—spurred the creation of the first custom-built aircraft for presidential use, a heavily modified VC-54 Skymaster. Though officially named “The Flying White House,” the new presidential aircraft became better known by its nickname, the “Sacred Cow.” President Harry Truman used the Sacred Cow as his presidential aircraft through much of his first term in office. In late 1947, the U.S. Air Force ordered a second custom-built presidential aircraft, a modified DC-6, which Truman named the Independence. While in office, Presidents Franklin D. Roosevelt and Harry Truman flew on a modified Douglas C-54, nicknamed the “Sacred Cow.”During Dwight D. Eisenhower’s two terms, the president flew on two different planes operated by the Air Force: the Columbine II, which was a customized, military version of Lockheed’s commercial airliner the Constellation, and the Columbine III, which was a Super Constellation. Embracing the jet age In the 1960s, the use of jet engine technology in U.S. commercial aircraft revolutionized air travel, allowing planes to fly higher, farther, and faster. Jet travel became associated with the glamorous and the elegant lifestyles of the “jet set” crowd. So it’s fitting that President John F. Kennedy—who was sometimes called the “the first celebrity president”—was the first White House occupant to fly in a jet, the Boeing 707. Kennedy’s aircraft was also the first painted in the distinctive light-blue-and-white scheme that’s still used today. First lady Jacqueline Kennedy developed it with the help of industrial designer Raymond Loewy. It would go on to serve eight presidents before leaving the presidential fleet in 1990, when Boeing delivered the first of two modified Boeing 747s. These are the aircraft that continue to serve as the president’s primary plane. Boeing signed a contract to provide two new aircraft in 2017, during Trump’s last term. In 2020, the company decided to refurbish two existing aircraft that were originally built for another customer. The refurbishment has been more cumbersome and expensive than building a new aircraft from scratch. But it’s the only option because Boeing closed its 747 assembly line in late 2022. A nickname sticks On a trip to Florida, the crew of Columbine II first used “Air Force One” as the plane’s call sign to clearly distinguish the plane from other air traffic. While the public has associated the name Air Force One with the modified Boeing 707s and 747s and their distinctive colors, any plane with the president aboard will carry that call sign. They include several smaller aircraft, also operated by the Air Force, such as the North American T-39 Sabreliner used to transport Lyndon B. Johnson to his ranch in Texas and the Lockheed VC-140B JetStars, the fleet of backup planes used by several presidents, which Johnson jokingly called “Air Force One Half.” A cultural and political symbol Air Force One has long served as a symbol of the power and prestige of the presidency. It became an indelible part of U.S. history in November 1963, when Johnson took his oath of office from Air Force One’s cabin while Kennedy’s body lay in rest in the back of the aircraft. Vice President Lyndon B. Johnson is sworn in as president aboard Air Force One following the assassination of President John F. Kennedy.Air Force One carried President Richard M. Nixon to China and the Soviet Union for historic diplomatic missions. But it also famously flew him from Andrews Air Force Base in Maryland to his home state, California, after he resigned from office. On that day, the plane took off as Air Force One. But it landed as SAM 27000, the plane’s call sign used when the president wasn’t on board. Trump has been compared to Nixon in more ways than one. And Trump’s complaint that Arab leaders have bigger and more impressive airplanes than the current Air Force One is reminiscent of Nixon’s own concerns of being outclassed on the world stage. When president, Nixon strongly advocated for American supersonic transport—a 270-passenger plane designed to be faster than the speed of sound—that he hoped could be modified to serve as a new Air Force One. He feared the failure to develop supersonic transport would relegate the U.S. to second-tier status, as other world leaders—particularly those from England, France, and the USSR—traversed the globe in sleeker, better performing aircraft. Trump’s concerns about Air Force One seem less focused on safety and security and more on size and opulence. His longing for a “palace in the sky” is befitting for a president drawn to soaring skyscrapers, lavish parades, and gold ornamentation. Janet Bednarek is a professor of history at the University of Dayton. This article is republished from The Conversation under a Creative Commons license. Read the original article. #trump #wants #air #force #one

Francis Kéré returned to his native Burkina Faso this month to celebrate the inauguration of his Berlin-based firm's latest design commission, the Thomas Sankara Mausoleum in Ouagadougou, the country's capital.  The new building commemorates Thomas Sankara, who ruled the West African nation as president from 1983 until his assassination on October 15, 1987, along with twelve of his government officials. Photo courtesy Kéré ArchitectureInside the mausoleum, the 13 tombs are arranged in a concentric circle, each with a single skylight placed above them. "At each hour, a new tomb is lit by the sun overhead, moving the visitor across the room and into a new space of remembrance," the architectural concept explains. Photo courtesy Kéré ArchitectureA large 112-foot dome covers the structure from the sun, while large louvred gates aligned with the direction of the prevailing east-west winds allow for passive ventilation of the building. Photo courtesy Kéré ArchitectureLocally-sourced late... #kérédesigned #thomas #sankara #mausoleum #memorializes

Every maker knows that familiar frustration: you’re deep into a project, carefully crafting something amazing, when your cutting tool snags, slips, or simply can’t handle the precision you need. Traditional cutting methods often leave rough edges, melted plastic, or splintered materials that require hours of additional sanding and finishing. Whether you’re trimming 3D prints, working with delicate materials, or adding intricate details to a model, the limitations of conventional tools can turn creative excitement into tedious damage control. The NeoBlade wireless ultrasonic cutter changes everything about how makers approach precision cutting tasks. This revolutionary tool harnesses the power of 40 kHz ultrasonic vibration technology in a compact, cordless package that weighs just 6.4 ounces with the battery installed. The high-frequency micro-vibrations allow the blade to slice through materials with minimal pressure and maximum control, creating the sensation of cutting through butter even when working with tough materials like carbon fiber, acrylics, or dense woods. This game-changing approach eliminates the struggle and frustration that often accompany detailed cutting work. Designer: HOZO Design Click Here to Buy Now:. Hurry, only 48/300 left! Raised over What truly sets the NeoBlade apart is its intelligent auto power adjustment system that intuitively adapts to different materials as you cut. This smart technology means you don’t need to constantly fiddle with settings or worry about applying too much force when transitioning between materials in mixed-media projects. The cutter automatically optimizes its performance whether you’re working with delicate PLA filament from a 3D print, tough ABS plastic, natural wood, or even PCB boards, allowing you to focus entirely on your creative vision rather than managing your tool. Dual cutting modes provide versatility for different project needs without compromising on performance. Precision mode delivers short, controlled cuts perfect for trimming intricate details, deburring edges, or making tiny adjustments to delicate parts. When you need to make longer cuts for splitting materials or creating grooves, Continuous mode keeps the blade vibrating without interruption, allowing smooth, consistent cutting action across the entire length of your workpiece. It’s a kind of flexibility that makes the NeoBlade suitable for everything from quick touch-ups to major fabrication tasks. Heat management has always been the Achilles’ heel of cutting tools, but the NeoBlade’s built-in turbofan cooling system solves this problem elegantly. The innovative ventilation design pulls air through side vents to maintain optimal operating temperature even during extended use. This cooling technology not only prevents the blade from overheating but also keeps the handle comfortable in your hand during long work sessions. More importantly, it ensures consistently clean cuts without the melted edges or material distortion that often occurs with overheated cutting tools. The versatility of the NeoBlade expands dramatically with its six interchangeable blade options, each designed for specific cutting challenges. The standard 30-degree edge handles most everyday tasks, while specialized options like the curved blade for fluid motions, the chisel-like blade for carving, and the double-edge blade for creating precise grooves allow you to tackle virtually any cutting task with confidence. Each blade comes in a spring-loaded dispenser with a one-way disposal slot, making replacements quick, easy, and safe without interrupting your creative flow. Changing the Battery Battery management becomes virtually effortless with the TurboDock charging system, eliminating the productivity-killing downtime that plagues other wireless tools. The dual-slot dock simultaneously charges both the NeoBlade and a spare battery in just 30 minutes, ensuring you always have power ready when inspiration strikes. This quick-swap battery system means you can work continuously on large projects without frustrating interruptions, maintaining your creative momentum from start to finish without being tethered to a power outlet or waiting for recharging. For serious makers who understand that the right tools transform not just the final product but the entire creative process, the NeoBlade represents a significant evolution in cutting technology. Its combination of wireless freedom, intelligent power management, and ultrasonic precision opens new possibilities for projects that previously seemed too delicate or complex to attempt. By removing the limitations and frustrations of traditional cutting methods, the NeoBlade doesn’t just help you make things; it fundamentally changes how you approach the making process, freeing your creativity to explore new materials, techniques, and designs without compromise. Click Here to Buy Now:. Hurry, only 48/300 left! Raised over The post NeoBlade: World’s Smallest Wireless Ultrasonic Cutter Revolutionizes DIY Projects first appeared on Yanko Design. #neoblade #worlds #smallest #wireless #ultrasonic

For many creatives, there are few times in life that are more thrilling and filled with possibilities than the summer you graduate. At the same time, there's usually a fair bit of anxiety, and in 2025, there are good reasons for that. If you're a creative graduate preparing to enter the workforce, you face a rapidly changing landscape shaped by technological advancements, shifting work paradigms and unprecedented challenges. In short, you'll be navigating uncharted waters as you transition from structured academic environments to the professional world. To help bridge this gap, we've gathered industry-leading advice from established leaders across various disciplines. These veterans have weathered economic shifts, tech revolutions, and cultural transformations throughout their careers. As such, their collective wisdom offers a compass for any graduate at a time of extraordinary opportunity and complexity. 1. Be kind to yourself First things first. You'll want to get into decent-paid work straight away, but in the current job market, that might not happen. That completely sucks, of course, but the important thing is that you don't beat yourself up over it. Creative director Charlie Bowden acknowledges the challenges of getting your foot in the door, even with a degree. "Converting an internship into a job is hard and may take longer than you'd like," he cautions. "So be kind to yourself. The best work comes when you're happy." Adrian Carroll, creative director at D8, agrees. "Don't be too hard on yourself if you don't land a role straight away," he stresses. "It's an incredibly tough market for graduates at the moment. Fewer people are setting up agencies, and more people are opting to freelance, which is tough if you don't have much experience." 2. Build your personal brand In a crowded jobs market, you need to stand out. That can be challenging for many creatives, especially if you're shy or introverted. But the important thing to understand, according to Simon Manchipp, founding partner at SomeOne, is this: "YOU ARE A BRAND. Like it or not." With that in mind, he recommends you: "Get your story straight. Make it stick. Website. Insta. LinkedIn. Behance. They should sing the same song—and it should sound like you. SHOW WORK. NOT PROMISES. Don't tell us what you want to do; show us what you did. Projects are better than platitudes. Side hustles count. So do failed ones. Especially failed ones." 3. Craft a portfolio with cut-through When it comes to your portfolio, Alex Rexworthy, co-founder and design director at Outlaw, advises showcasing development alongside final work. "Show your process, not just a flashy render," he urges. "Take the reader on a journey through each project in your portfolio. Start with the brief, share your initial ideas, and explain how you arrived at the final result." Above all, encourages Cat How, founder and ECD at How Studio, make an impact. "Go big, or go home," she says. "This is your time to shine, so you owe it to yourself to be OUTRAGEOUS." 4. Stay true to your vision When you're struggling to get work, it's natural to want to apply for everything and maybe even pretend to be someone you're not. But Paul Leon, creative director at U037, reckons you'll ultimately be more successful if you focus on your authentic creative path. "Focus on what you really want to do and head for that," he advises. "If you feel something's not right for you, don't be afraid to speak up, and don't be afraid to say no. Value yourself: the journey is where all the good stuff happens. Detours, etc, are fine." When applying for a job at an agency, you obviously need to do your research. But Charlie notes that you need to look beyond an agency's portfolio. "Consider the culture, not just the work," he explains. "Certain agencies make great work, but their culture may not be right for you. Internships are a great way of getting a gauge of this, but so is asking to come in for coffee and a crit, so you can see what the agency is like first-hand." 5. Embrace the beginner's mindset You need confidence to succeed in life, but that confidence shouldn't spill over into overconfidence or the idea that you already know it all; you don't. With that in mind, Kiser Barnes, partner and chief creative officer at Red Antler, stresses the importance of humility paired with curiosity. "Be humble and stay wildly curious," he says. "Finishing school feels like reaching the peak… until you realise you're right back at the bottom of a new hill again. That's okay. The people who thrive in this industry are the ones most open to learning, evolving and collaborating." Barrington Reeves, creative director at Too Gallus, offers similar advice. "Be ready to learn," he urges. "The change from education to commercial can sometimes be daunting. But although you've just been through a full course of education, the real learning begins now. " Part of that means adapting to commercial timeframes. "Don't be discouraged by the pace of change and iteration," he stresses. "While university might teach you to nurture and grow an idea over a prolonged period of time, in the commercial market, it is an incredible skill to be able to think—and fail—fast." 6. Develop leadership skills early Being humble doesn't mean you can't have one eye on the future and start developing your leadership skills. With that in mind, Ashleigh Hansberger, co-founder and COO at Motto, suggests you: "Start studying business and leadership now. Leadership isn't about title; it's about behaviour. That means building discipline, making clear decisions, being accountable, and taking responsibility for your actions and impact. "Practise your own leadership development early," she adds. "Read about it. Reflect on it. Try it. Accumulate experience and wisdom through the unglamorous work of getting 1% better every day. That's how you grow into someone people want to follow." 7. Master your craft while embracing technology Emerging technologies might be disrupting everything right now, with negative consequences for employment. But Kiser encourages graduates to engage with them all the same. "Don't be afraid of AI—tinker with it, play with it, understand it," he says. "Use your fresh perspective to imagine better ways of working. The world doesn't need more copies of what already exists. It needs people who can push the whole system forward." Matthew Schneider, director of product marketing at LucidLink, takes a similar line. "Know your craft, but don't be afraid of new tech," he advises. "When I started, a single video editing system cost well over k; now, a subscription to professional editing software might cost as little as monthly." And the moral of this story? "Now people have access to great tools, what sets you apart is how you use them and the story you ultimately tell with them," Matthew explains. "Growing up an aspiring filmmaker myself, my mother would always tell me, 'It's not the story; it's the storytelling'. She's right!" Tom Munckton, ECD at Fold7Design, similarly notes the accelerating pace of industry evolution right now. "The industry used to shift annually; now it feels more like monthly—with processes and project delivery being challenged by AI and other democratising factors," he says. Ultimately, though, he recommends focusing on creative direction over specific tools. "Regardless of the tool or method you use right now or in the future, confidence will come from being clear about the type of work you want to make and making constant steps towards that," he reasons. 8. Embrace discomfort as growth Finding all this stressful and frightening? Many creative people feel like they don't "belong" in normal society, but the good news is that agencies are generally pretty friendly places and are probably filled with other people who've felt like outsiders themselves. Take Claire Parker, group creative partner at The Chase. "I never fitted in at school—couldn't do maths, couldn't spell—but I could create," she recalls. "The world needs curious minds like yours. Creativity isn't about having all the answers; it's about asking better questions. "Stay interested, stay committed, and back yourself even when you feel out of your depth," she adds. "In fact, especially then, as David Bowie said, 'Always go a little further into the water'. That's where the magic happens. Be more Bowie." #creative #leaders #share #their #golden

Are you an experienced engineer looking for an exciting start-up opportunity in the process of finding PMF?Keep reading!We’re Koast, on a mission to revolutionize Meta media buying with AI-powered workflows. We’re looking for a talented Senior Full-Stack Engineer to join our growing engineering team. As the third member of our technical team, you’ll take full ownership of core systems, from data models to production deployments. You’ll play a crucial role in building features that power sophisticated ad campaigns reaching millions of people globally.What You’ll Do DailyBackend MasteryArchitect, implement, and operate our Node.js + TypeScript services - data models through CI/CD - without hand-holding.Front-End CraftsmanshipShip React components using hooks and state libs, nail pixel-perfect UIs, and squash cross-browser quirks.LLM Feature IntegrationDesign prompts, manage streaming, and build rate-limit/cost controls around GPT-class models.End to End Feature LifecycleTake features from spec → prototype → code → deploy → monitor, owning each step without hand-holding.Must-Have Technical ExperienceNode.js & TypeScript— you’ve owned production services from spec to scale.React— modern hooks, state libs, and deep browser-debug skills.End-to-End Product Mindset — you think in features, not tickets, and drive them all the way home.LLM-Powered Features — hands-on prompt design, streaming integrations, and cost/rate-limit handling.AI-Augmented Dev — Copilot/Cursorare your daily co-pilots.Nice-to-HaveNestJS or another opinionated Node frameworkNext.js / React Server ComponentsDev-ex tooling: GitHub Actions, automated testing, observability, profilingDesign-system fluency: Tailwind, MUI, Chakra, or custom component libsHow You WorkWe live by these values - if they sound like you, you’ll fit right in....Mastery: You push beyond “good enough,” constantly honing your craft.Push the Boundaries: You tackle ambitious challenges and unlock new possibilities.Embrace the Challenge: You collaborate shoulder-to-shoulder when the stakes are high.Responsibility: You own your commitments, deliver on time, and keep teammates aligned.Dedication: You dive head-first into tough problems, iterating until they’re solved.Collaboration: You communicate clearly, share knowledge freely, and build together.Why Koast.ai?Early-Stage ImpactYour work will define our architecture and ship features that 100+ customers rely on every day.AI & Ads at ScaleJoin the frontier of AI-powered media buying - supercharge budgets and decision-making in real time for some of the best ad teams in the world.Growth & OwnershipStart as a contractor, grow into a full-time role, and potentially earn equity.Remote & FlexibleWork from anywhere on your schedule—just deliver high-quality code.🚨 Bonus: Send a 1–2 min video introducing yourself and why this role excites you—we’ll bump you to the top of the stack!Ready to apply?Email your resume, GitHub/portfolio links, andthat intro video towith the subject line “Koast to the moon.” In your email, we’d love to hear how you’ll make an impact here that no “regular” engineer could. That’s it!Can’t wait to build the future of autonomous ad buying with you!Best,The Koast.ai TeamApply NowLet's start your dream job Apply now Meet JobCopilot: Your Personal AI Job HunterAutomatically Apply to Remote Full-Stack Programming JobsJust set your preferences and Job Copilot will do the rest-finding, filtering, and applying while you focus on what matters. Activate JobCopilot #koastai #senior #fullstack #engineer #exciting

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Microsoft launched a new enterprise platform that harnesses artificial intelligence to dramatically accelerate scientific research and development, potentially compressing years of laboratory work into weeks or even days. The platform, called Microsoft Discovery, leverages specialized AI agents and high-performance computing to help scientists and engineers tackle complex research challenges without requiring them to write code, the company announced Monday at its annual Build developer conference. “What we’re doing is really taking a look at how we can apply advancements in agentic AI and compute work, and then on to quantum computing, and apply it in the really important space, which is science,” said Jason Zander, Corporate Vice President of Strategic Missions and Technologies at Microsoft, in an exclusive interview with VentureBeat. The system has already demonstrated its potential in Microsoft’s own research, where it helped discover a novel coolant for immersion cooling of data centers in approximately 200 hours — a process that traditionally would have taken months or years. “In 200 hours with this framework, we were able to go through and screen 367,000 potential candidates that we came up with,” Zander explained. “We actually took it to a partner, and they actually synthesized it.” How Microsoft is putting supercomputing power in the hands of everyday scientists Microsoft Discovery represents a significant step toward democratizing advanced scientific tools, allowing researchers to interact with supercomputers and complex simulations using natural language rather than requiring specialized programming skills. “It’s about empowering scientists to transform the entire discovery process with agentic AI,” Zander emphasized. “My PhD is in biology. I’m not a computer scientist, but if you can unlock that power of a supercomputer just by allowing me to prompt it, that’s very powerful.” The platform addresses a key challenge in scientific research: the disconnect between domain expertise and computational skills. Traditionally, scientists would need to learn programming to leverage advanced computing tools, creating a bottleneck in the research process. This democratization could prove particularly valuable for smaller research institutions that lack the resources to hire computational specialists to augment their scientific teams. By allowing domain experts to directly query complex simulations and run experiments through natural language, Microsoft is effectively lowering the barrier to entry for cutting-edge research techniques. “As a scientist, I’m a biologist. I don’t know how to write computer code. I don’t want to spend all my time going into an editor and writing scripts and stuff to ask a supercomputer to do something,” Zander said. “I just wanted, like, this is what I want in plain English or plain language, and go do it.” Inside Microsoft Discovery: AI ‘postdocs’ that can screen hundreds of thousands of experiments Microsoft Discovery operates through what Zander described as a team of AI “postdocs” — specialized agents that can perform different aspects of the scientific process, from literature review to computational simulations. “These postdoc agents do that work,” Zander explained. “It’s like having a team of folks that just got their PhD. They’re like residents in medicine — you’re in the hospital, but you’re still finishing.” The platform combines two key components: foundational models that handle planning and specialized models trained for particular scientific domains like physics, chemistry, and biology. What makes this approach unique is how it blends general AI capabilities with deeply specialized scientific knowledge. “The core process, you’ll find two parts of this,” Zander said. “One is we’re using foundational models for doing the planning. The other piece is, on the AI side, a set of models that are designed specifically for particular domains of science, that includes physics, chemistry, biology.” According to a company statement, Microsoft Discovery is built on a “graph-based knowledge engine” that constructs nuanced relationships between proprietary data and external scientific research. This allows it to understand conflicting theories and diverse experimental results across disciplines, while maintaining transparency by tracking sources and reasoning processes. At the center of the user experience is a Copilot interface that orchestrates these specialized agents based on researcher prompts, identifying which agents to leverage and setting up end-to-end workflows. This interface essentially acts as the central hub where human scientists can guide their virtual research team. From months to hours: How Microsoft used its own AI to solve a critical data center cooling challenge To demonstrate the platform’s capabilities, Microsoft used Microsoft Discovery to address a pressing challenge in data center technology: finding alternatives to coolants containing PFAS, so-called “forever chemicals” that are increasingly facing regulatory restrictions. Current data center cooling methods often rely on harmful chemicals that are becoming untenable as global regulations push to ban these substances. Microsoft researchers used the platform to screen hundreds of thousands of potential alternatives. “We did prototypes on this. Actually, when I owned Azure, I did a prototype eight years ago, and it works super well, actually,” Zander said. “It’s actually like 60 to 90% more efficient than just air cooling. The big problem is that coolant material that’s on market has PFAS in it.” After identifying promising candidates, Microsoft synthesized the coolant and demonstrated it cooling a GPU running a video game. While this specific application remains experimental, it illustrates how Microsoft Discovery can compress development timelines for companies facing regulatory challenges. The implications extend far beyond Microsoft’s own data centers. Any industry facing similar regulatory pressure to replace established chemicals or materials could potentially use this approach to accelerate their R&D cycles dramatically. What once would have been multi-year development processes might now be completed in a matter of months. Daniel Pope, founder of Submer, a company focused on sustainable data centers, was quoted in the press release saying: “The speed and depth of molecular screening achieved by Microsoft Discovery would’ve been impossible with traditional methods. What once took years of lab work and trial-and-error, Microsoft Discovery can accomplish in just weeks, and with greater confidence.” Pharma, beauty, and chips: The major companies already lining up to use Microsoft’s new scientific AI Microsoft is building an ecosystem of partners across diverse industries to implement the platform, indicating its broad applicability beyond the company’s internal research needs. Pharmaceutical giant GSK is exploring the platform for its potential to transform medicinal chemistry. The company stated an intent to partner with Microsoft to advance “GSK’s generative platforms for parallel prediction and testing, creating new medicines with greater speed and precision.” In the consumer space, Estée Lauder plans to harness Microsoft Discovery to accelerate product development in skincare, makeup, and fragrance. “The Microsoft Discovery platform will help us to unleash the power of our data to drive fast, agile, breakthrough innovation and high-quality, personalized products that will delight our consumers,” said Kosmas Kretsos, PhD, MBA, Vice President of R&D and Innovation Technology at Estée Lauder Companies. Microsoft is also expanding its partnership with Nvidia to integrate Nvidia’s ALCHEMI and BioNeMo NIM microservices with Microsoft Discovery, enabling faster breakthroughs in materials and life sciences. This partnership will allow researchers to leverage state-of-the-art inference capabilities for candidate identification, property mapping, and synthetic data generation. “AI is dramatically accelerating the pace of scientific discovery,” said Dion Harris, senior director of accelerated data center solutions at Nvidia. “By integrating Nvidia ALCHEMI and BioNeMo NIM microservices into Azure Discovery, we’re giving scientists the ability to move from data to discovery with unprecedented speed, scale, and efficiency.” In the semiconductor space, Microsoft plans to integrate Synopsys’ industry solutions to accelerate chip design and development. Sassine Ghazi, President and CEO of Synopsys, described semiconductor engineering as “among the most complex, consequential and high-stakes scientific endeavors of our time,” making it “an extremely compelling use case for artificial intelligence.” System integrators Accenture and Capgemini will help customers implement and scale Microsoft Discovery deployments, bridging the gap between Microsoft’s technology and industry-specific applications. Microsoft’s quantum strategy: Why Discovery is just the beginning of a scientific computing revolution Microsoft Discovery also represents a stepping stone toward the company’s broader quantum computing ambitions. Zander explained that while the platform currently uses conventional high-performance computing, it’s designed with future quantum capabilities in mind. “Science is a hero scenario for a quantum computer,” Zander said. “If you ask yourself, what can a quantum computer do? It’s extremely good at exploring complicated problem spaces that classic computers just aren’t able to do.” Microsoft recently announced advancements in quantum computing with its Majorana one chip, which the company claims could potentially fit a million qubits “in the palm of your hand” — compared to competing approaches that might require “a football field worth of equipment.” “General generative chemistry — we think the hero scenario for high-scale quantum computers is actually chemistry,” Zander explained. “Because what it can do is take a small amount of data and explore a space that would take millions of years for a classic, even the largest supercomputer, to do.” This connection between today’s AI-driven discovery platform and tomorrow’s quantum computers reveals Microsoft’s long-term strategy: building the software infrastructure and user experience today that will eventually harness the revolutionary capabilities of quantum computing when the hardware matures. Zander envisions a future where quantum computers design their own successors: “One of the first things that I want to do when I get the quantum computer that does that kind of work is I’m going to go give it my material stack for my chip. I’m going to basically say, ‘Okay, go simulate that sucker. Tell me how I build a new, a better, new version of you.'” Guarding against misuse: The ethical guardrails Microsoft built into its scientific platform With the powerful capabilities Microsoft Discovery offers, questions about potential misuse naturally arise. Zander emphasized that the platform incorporates Microsoft’s responsible AI framework. “We have the responsible AI program, and it’s been around, actually I think we were one of the first companies to actually put that kind of framework into place,” Zander said. “Discovery absolutely is following all responsible AI guidelines.” These safeguards include ethical use guidelines and content moderation similar to those implemented in consumer AI systems, but tailored for scientific applications. The company appears to be taking a proactive approach to identifying potential misuse scenarios. “We already look for particular types of algorithms that could be harmful and try and flag those in content moderation style,” Zander explained. “Again, the analogy would be very similar to what a consumer kind of bot would do.” This focus on responsible innovation reflects the dual-use nature of powerful scientific tools — the same platform that could accelerate lifesaving drug discovery could potentially be misused in other contexts. Microsoft’s approach attempts to balance innovation with appropriate safeguards, though the effectiveness of these measures will only become clear as the platform is adopted more widely. The bigger picture: How Microsoft’s AI platform could reshape the pace of human innovation Microsoft’s entry into scientific AI comes at a time when the field of accelerated discovery is heating up. The ability to compress research timelines could have profound implications for addressing urgent global challenges, from drug discovery to climate change solutions. What differentiates Microsoft’s approach is its focus on accessibility for non-computational scientists and its integration with the company’s existing cloud infrastructure and future quantum ambitions. By allowing domain experts to directly leverage advanced computing without intermediaries, Microsoft could potentially remove a significant bottleneck in scientific progress. “The big efficiencies are coming from places where, instead of me cramming additional domain knowledge, in this case, a scientist having learned to code, we’re basically saying, ‘Actually, we’ll let the genetic AI do that, you can do what you do, which is use your PhD and get forward progress,'” Zander explained. This democratization of advanced computational methods could lead to a fundamental shift in how scientific research is conducted globally. Smaller labs and institutions in regions with less computational infrastructure might suddenly gain access to capabilities previously available only to elite research institutions. However, the success of Microsoft Discovery will ultimately depend on how effectively it integrates into complex existing research workflows and whether its AI agents can truly understand the nuances of specialized scientific domains. The scientific community is notoriously rigorous and skeptical of new methodologies – Microsoft will need to demonstrate consistent, reproducible results to gain widespread adoption. The platform enters private preview today, with pricing details yet to be announced. Microsoft indicates that smaller research labs will be able to access the platform through Azure, with costs structured similarly to other cloud services. “At the end of the day, our goal, from a business perspective, is that it’s all about enabling that core platform, as opposed to you having to stand up,” Zander said. “It’ll just basically ride on top of the cloud and make it much easier for people to do.” Accelerating the future: When AI meets scientific method As Microsoft builds out its ambitious scientific AI platform, it positions itself at a unique juncture in the history of both computing and scientific discovery. The scientific method – a process refined over centuries – is now being augmented by some of the most advanced artificial intelligence ever created. Microsoft Discovery represents a bet that the next era of scientific breakthroughs won’t come from either brilliant human minds or powerful AI systems working in isolation, but from their collaboration – where AI handles the computational heavy lifting while human scientists provide the creativity, intuition, and critical thinking that machines still lack. “If you think about chemistry, materials sciences, materials actually impact about 98% of the world,” Zander noted. “Everything, the desks, the displays we’re using, the clothing that we’re wearing. It’s all materials.” The implications of accelerating discovery in these domains extend far beyond Microsoft’s business interests or even the tech industry. If successful, platforms like Microsoft Discovery could fundamentally alter the pace at which humanity can innovate in response to existential challenges – from climate change to pandemic prevention. The question now isn’t whether AI will transform scientific research, but how quickly and how deeply. As Zander put it: “We need to start working faster.” In a world facing increasingly complex challenges, Microsoft is betting that the combination of human scientific expertise and agentic AI might be exactly the acceleration we need. Daily insights on business use cases with VB Daily If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI. Read our Privacy Policy Thanks for subscribing. Check out more VB newsletters here. An error occured. #microsoft #just #launched #that #discovered