pytorch
(11)
Dmytro Dzhulgakov Learned to Race Code. Now He Races AI Models.
Before he worked on the machinery behind generative AI, Dmytro Dzhulgakov was a Kharkiv student who wrote code for an Intel 8080 and competed against the world’s fastest programmers. At Fireworks AI, the contest has changed, but the clock still matters.
Lin Qiao Builds AI Infrastructure Like Her Father Built Ships
Before she led PyTorch and founded Fireworks AI, Lin Qiao learned to read cargo-ship blueprints beside her father. Her career has followed the same discipline ever since: absorb the complexity, expose the useful parts, and build for people who need the system to hold.
Inferact
Inferact is a Berkeley- and San Francisco-based AI infrastructure company founded by the creators and core maintainers of vLLM, the most widely used open-source engine for running large language models. The company commercializes vLLM - which at any moment powers inference on more than 400,000 GPUs worldwide - by funneling paid engineering resources back into the open project while building a next-generation commercial 'universal inference layer' meant to make AI inference cheaper and faster. Inferact launched publicly in January 2026 with a $150 million seed round at an $800 million valuation, co-led by Andreessen Horowitz and Lightspeed.
Arago
Arago is a Paris and Palo Alto deeptech startup building a photonic AI accelerator that uses light instead of electrons to run AI workloads, targeting 10x lower energy consumption than today's leading GPUs at equivalent performance and cost while staying compatible with PyTorch and standard hardware ecosystems.
Gimlet Labs, Inc.
Gimlet Labs is a San Francisco applied research company building the first multi-silicon inference cloud - software that runs AI workloads simultaneously across CPUs, GPUs, and specialized accelerators. Founded by ex-Google and ex-NVIDIA engineers behind the Pixie observability project, Gimlet ships Gimlet Cloud (serverless agent inference) and kforge (autonomous kernel generation from PyTorch). The company exited stealth in October 2025 with eight-figure revenue and raised an $80M Series A led by Menlo Ventures in March 2026.
Meta
Meta Platforms is the company behind Facebook, Instagram, WhatsApp, Messenger, and Threads - a Family of Apps used by more than 3 billion people daily. Founded as Facebook in 2004 and renamed Meta in 2021, it runs one of the world's largest digital advertising businesses while pouring tens of billions into artificial intelligence (the open-weight Llama models and Meta AI assistant) and wearable computing (Quest headsets and Ray-Ban Meta smart glasses). In 2025 it reported roughly $201 billion in revenue.
Sebastian Raschka
Sebastian Raschka is a German-born AI/ML researcher, educator, and author who has built one of the most trusted independent voices in the machine learning community. Through his Substack newsletter 'Ahead of AI' (184,000+ subscribers), bestselling books like 'Build a Large Language Model (From Scratch)', and 91,000+ starred GitHub repositories, he demystifies cutting-edge AI for practitioners worldwide. After a stint as an Assistant Professor at UW-Madison and a role as Staff Research Engineer at Lightning AI, he now runs RAIR Lab as an independent researcher, writer, and consultant.
Daniel Bourke
Daniel Bourke is an Australian machine learning engineer, educator, and content creator who teaches over 230,000 students worldwide through the Zero to Mastery Academy. Operating under the brand 'mrdbourke', he built a self-designed AI education path from a film degree and zero coding experience, and now creates courses on PyTorch, TensorFlow, and Hugging Face. He co-founded Nutrify, an AI-powered food tracking app, with his brother Joshua, and published his debut novel 'Charlie Walks' in 2024. His newsletter 'Eat, Move, Learn, Make' blends personal philosophy with technical insight.
Tim Dettmers
Tim Dettmers is an Assistant Professor at Carnegie Mellon University and Research Scientist at the Allen Institute for AI (AI2), best known for making large language models accessible on consumer hardware. He created the bitsandbytes library (2.2M monthly installs), co-authored QLoRA - a technique enabling fine-tuning of 65B-parameter models on a single GPU - and pioneered LLM.int8() quantization. With over 18,000 citations across his work, Dettmers has become one of the most influential voices in efficient deep learning, consistently arguing that computational democratization - not AGI hype - is where the real progress lives.
Vik Paruchuri
Vik Paruchuri is the founder and CEO of Datalab, an AI startup building small, efficient foundation models for document intelligence. A self-taught ML engineer who majored in American History, he previously founded Dataquest - an online learning platform that taught data skills to over 1 million students. His open-source projects (Marker, Surya, Chandra OCR) have earned thousands of GitHub stars and benchmark-topping accuracy scores. He publishes 'The Vik Letter' newsletter covering semiconductors and tech.
Fireworks AI
Fireworks AI is a generative AI inference platform founded in 2022 by seven engineers — five of whom built PyTorch at Meta — that gives enterprises fast, cost-efficient, and customizable access to hundreds of open-source models. The company's proprietary FireAttention kernels and speculative-execution engine deliver up to 40× faster inference and 8× cost reduction versus alternatives, while its fine-tuning and model-deployment tooling lets companies own their AI stack end-to-end. With $327M+ raised, a $4B valuation, 10,000+ customers including Samsung, Uber, Shopify, and Cursor, and a $315M annualized run-rate as of early 2026, Fireworks AI has become the go-to inference layer for production generative AI applications.