{"id":101365,"title":"PerfectBit: AI Training data, correct by construction","tagline":"Specialized data for the world's most powerful models.","body":"# **TL;DR**\n\n_PerfectBit creates high-quality training data that's correct by construction. We verify against physics simulators, scientific databases, formal proof systems. It's what models need to cure hallucination, close common-sense gaps, and reach superintelligence. Reinforcement Learning from Verified Rewards (RLVR) was an early step in this direction. We go further. LLMs, robotics, AI for Science, and more._\n\n\u003chttps://youtu.be/RrFKfbO6XlY\u003e\n\n# **The problem: the failures standing between today's models and superintelligence**\n\nFrontier labs are bottlenecked on four data problems:\n\n* **Hallucination.** Models trained on noisy web scrapes produce confident, fluent statements that are wrong.\n* **Common-sense failure.** Models that solve graduate-level math still trip on trivial problems for a five-year-old.\n* **Information scarcity.** Scientific facts, formal derivations, peer-reviewed findings — they're a thin minority of tokens in any open-web scrape, drowned out by orders of magnitude of recipes, marketing copy, forum chatter, and AI-generated slop. Scaling the scrape doesn't change the ratio.\n* **Human annotation doesn't scale.** It's slow, expensive, and capped at human ability — superintelligence won't come from labeling at human throughput.\n\n# **What we do**\n\n* PerfectBit creates high-quality training data that's correct by construction. Every sample is generated against an oracle — physics simulator, scientific database, formal proof system — and verified before it ships, in text, code, image, audio, or video. For LLMs, we project observations about the world into natural language; for verticals like robotics and AI-for-science, we ship purpose-built datasets.\n\n# **Why us**\n\nWe've trained the models we're now building data for.\n\n* **Peter Vajda** — Former **Director of Media Generation at Meta**. Eleven years at Meta. Most recently led Media GenAI foundation-model R\u0026amp;D: **Emu** (text-to-image), image editing, **Movie Gen**, text-to-video, video editing, character-consistent generation. Earlier led efficient deep learning for computer vision powering on-device Augmented Reality and Virtual Reality (AR/VR) models. **Assistant Visiting Professor at Stanford** before Meta. PhD in Computer Science.\n* **Seiji Yamamoto** — Led teams in the **Core Llama group at Meta Superintelligence Labs**. Nine years at Meta as a **Senior Staff Research Scientist** across LLM pre- and post-training, inference optimization, full-duplex speech, and computer vision. **PhD in Physics**, publications in **Proceedings of the National Academy of Sciences** and **Physical Review Letters** (co-authored with a Fields Medalist). Stanford, Rice, Columbia; postdoc at a National Lab.\n\nBetween us: foundation-model training across text, image, video, and speech, shipped to billions of users — plus scientific chops to design the verifier stacks.\n\n# **We’d love to talk more**\n\nWe're talking to a small number of frontier AI labs about pilot engagements.\n\n* **Heads of data / data partnerships** at frontier labs\n* **Research leads** running pre-training, mid-training, post-training, or reasoning programs and shopping for a corpus that doesn't exist yet\n* **Model-training teams** preparing a next-gen training run who want a correct-by-construction supplement in the mix\n* **Researchers in formal methods, scientific computing, or domain-specific simulators** who want to contribute a verifier stack — we'd love to talk\n\nWe're also **hiring research scientists and engineers in San Francisco** (in-person). If you've shipped foundation-model training, built formal/simulation tooling or spent time developing AI training data at production scale, please reach out.\n\n# **Contact**\n\n* **Web:** [**https://perfectbit.ai**](https://perfectbit.ai)\n* **Email:** [**founders@perfectbit.ai**](mailto:founders@perfectbit.ai)\n* **Founders:** Peter Vajda, Seiji Yamamoto","slug":"QMv-perfectbit-ai-training-data-correct-by-construction","created_at":"2026-05-14T19:03:28.026Z","updated_at":"2026-09-19T01:40:19.814Z","total_vote_count":8,"url":"https://www.ycombinator.com/launches/QMv-perfectbit-ai-training-data-correct-by-construction","share_image_url":"//bookface-static.ycombinator.com/assets/ycdc/yc-og-image-c440a0ad1dacfb86eeeb343717479cc54d256614449b4ef719977a0a451f8bc8.png","company":{"id":31333,"name":"PerfectBit, Inc.","slug":"perfectbit-inc","url":"https://perfectbit.ai","logo":"https://bookface-images.s3.amazonaws.com/small_logos/900c210e3f57ee42221dbb5ea21f81418b553e70.png","batch":"Spring 2026","industry":"Industrials","tags":["Artificial Intelligence","Machine Learning","Robotics"],"search_path":"https://bookface.ycombinator.com/company/31333"}}