Sepal AI
Sepal is a data research company on a mission to advance human knowledge and capabilities through safe AI. We partner with the world’s leading AI labs and enterprises to help their models get better at the tasks people actually want them to do. We’ve built a Cloud-Native Agent Dataset Factory which turns the process of generating evaluation and training data from manual, inconsistent, and labor-intensive into something automated, standardized, and scalable. Sepal AI was founded in 2024 by engine
Y Combinator
yc-company-directory #2911Sepal is a data research company on a mission to advance human knowledge and capabilities through safe AI. We partner with the world’s leading AI labs and enterprises to help their models get better at the tasks people actually want them to do. We’ve built a Cloud-Native Agent Dataset Factory which turns the process of generating evaluation and training data from manual, inconsistent, and labor-intensive into something automated, standardized, and scalable. Sepal AI was founded in 2024 by engine
sourceSepal AI is classified as AI agents from source descriptions and fund-directory context. Similar companies below are ranked from offline product facets and meaningful description overlap; category and source affiliation can only strengthen an existing product match. For lookup-created rows, the profile starts as pending enrichment until deeper source collection runs.
Local clusters rerank extracted product concepts and description terms. The generated graph remains the fallback for sparse descriptions: offline reciprocal product-similarity graph: extracted concepts + description terms + bounded category/affiliation boosts. Minimum fallback score: 0; max competitors: 6.
Companies similar to Sepal AI
A deterministic local cluster built from offline product facets and meaningful description terms. Category and selected-institution affiliation provide bounded tie-breaks. Open any peer to continue exploring its cluster.
Halluminate
match 80We help foundation model labs and enterprises train computer use AI with data and sandboxes. A knowledge worker spends 90% of their day on a computer. A productive AI agent must similarly learn how to use our computers, browsers, and software to deliver real world value. Today, browser and computer use AI is inaccurate, slow, and expensive. Model labs and enterprises who want to improve these agents are bottlenecked by two resources: 1) High quality datasets for benchmarking and evaluations 2) R
shared extracted concepts: enterprises; shared product terms: model, labs, enterprise, train; same category: AI agents; shared affiliation: Y Combinator
Abundant
match 65Hello! 👋 We are a team of former ML engineers, founders, roboticists and ops leads who obsess about data and its impact on safe, reliable AI. We specialize in creating environments and datasets for RL by leveraging our experience in simulation and model training. By the numbers: • Powering 3 of the top 6 global AI labs and multiple Fortune 500 enterprises • Billions of training tokens generated each month, 2x month over month • Exclusive, global network of over 500 domain experts We believe hum
shared product terms: lead, safe, dataset, model; shared affiliation: Y Combinator
hud
match 57HUD (YC W25) is developing agentic evals and RL environments for Computer Use Agents (CUAs) that browse the web for frontier AI labs. Our CUA Evals framework is the first comprehensive evaluation tool for CUAs. People don't actually know if AI agents are working reliably. To make AI agents work in the real world, we need detailed evals for a huge range of tasks. We're backed by Y Combinator, and work closely with frontier AI labs to provide agent evaluation and training infrastructure at scale.
shared product terms: labs, evaluation, people, actually; shared product theme: developer infrastructure; same category: AI agents; shared affiliation: Y Combinator
Aviro
match 57Aviro is a research partner with four frontier AI labs, F500 companies, and some of the top RL data vendors building post-training datasets. We focus on tasks involving thousands of tool steps for coding, computer use, and knowledge work.
shared product terms: research, partner, labs, train; shared affiliation: Y Combinator
Asimov
match 57Asimov collects real-world human movement data from households and businesses to train humanoid robots. Unlike factory datasets that capture the same tasks in the same environments, we deliver the full diversity of real human environments, thousands of hours a day to leading labs.
shared product terms: human, train, factory, dataset; shared affiliation: Y Combinator
Lucidic AI
match 52Lucidic emulates model training without changing model weights. The last decade made models intelligent, but intelligence is not the same as experience. Humans do not get better by memorizing thousands of examples; we build learning systems around ourselves: skills, memories, critique, practice, tools, and guidance. And every person learns differently because every task is different. Lucidic brings that idea to AI agents by training a custom learning system for each agent, so it learns what to r
shared product terms: model, train, human, do; same category: AI agents; shared affiliation: Y Combinator