RL environments and datasets for software engineering; creator of SWE-Marathon.
1 listed itemJesse Hu, Meji Abidoye, Ke Huang
Waymo·Google·Meta·AWS ML and data-operations background
Company size 1–10Pack, price, samples & audit · unlock after NDA
RL ENVIRONMENT SOURCING FOR AI LABS
Find the right environments.
Start with a sample.
Narrow by domain and environment. Compare suppliers.
Write your requirements once. Draft a request for each supplier.
72 suppliers
Custom tool-use, desktop and browser environments with expert rubrics and verifiers.
3 listed itemsSpencer Mateega, Carlos Georgescu
Penn·UBC · internships at Meta, Google, Citadel, Silver Lake, Morgan Stanley
Company size 51–200Pack, price, samples & audit · unlock after NDA
Builds proprietary robot evaluation environments and custom facility simulations. RL training rights and environment delivery terms remain unverified; request a demo to confirm scope.
2 listed itemsTeam not disclosed
Team described as ex-OpenAI/xAI (individual backgrounds unverified)
Company size 1–10Pack, price, samples & audit · unlock after NDA
RL simulation environments for tool-use and knowledge-work agents, with hidden-verifier scoring and Harbor-format tasks.
2 listed itemsSpandana Govindgari, Navi Singh, Girish Kumar
Co-founders with ex-Meta, Apple and MIT research backgrounds
Company size 1–10Pack, price, samples & audit · unlock after NDA
Programmatically generated environments, tasks and verifiers for long-horizon work; private beta.
1 listed itemAlexandre Sallinen, Guillaume Allegre
Company size 1–10Pack, price, samples & audit · unlock after NDA
Custom RLHF preference data, post-training datasets and expert evaluation, built to your use case.
2 listed itemsPratik Karki, Mannat Sandhu
Company size 1–10Pack, price, samples & audit · unlock after NDA
RL tasks and verifiers for software, finance, healthcare and enterprise work.
9 listed itemsJulie Vonwiller, Chris Vonwiller
CEO Ryan Kolln: telecom/data engineer, NYU MBA, BCG strategy consulting; joined Appen in 2018
Company size 500+Pack, price, samples & audit · unlock after NDA
RL environments reflecting real finance, healthcare and legal workflows.
1 listed itemChirag Modi, Armin Schöpf
Medical, finance and legal experts working with AI engineers
Company size 11–50Pack, price, samples & audit · unlock after NDA
Verifiable environments for cybersecurity, SRE, compilation and STEM reasoning.
5 listed itemsMykyta Mudryi, Markiian Chaklosh
Founder: security consulting at Leviathan Security Group
Company size 11–50Pack, price, samples & audit · unlock after NDA
Long-horizon coding, computer use, professional work and autonomous research environments.
4 listed itemsAarush Kukreja, Ayush Karupakula
Aarush Kukreja: Princeton CS · SWE-bench Verified / SWE-agent research at PLI · MIT·Caltech research (per his own bio)
Company size 1–10Pack, price, samples & audit · unlock after NDA
Company-scale simulations for training agents on enterprise work.
1 listed itemMahesh Sathiamoorthy, Alex Dimakis
Google DeepMind background · UC Berkeley faculty
Company size 11–50Pack, price, samples & audit · unlock after NDA
Savant RL Environments offers five training-objective configurations within its security environment platform. Pilot access is available by inquiry.
5 listed itemsCasey Ellis, Chris Raethke, Sergei Belokamen
Dave Gerry: WhiteHat Security CRO/COO, Veracode · David Brumley: Mayhem founder, Carnegie Mellon professor, led the DARPA CGC-winning team
Company size 201–500Pack, price, samples & audit · unlock after NDA
24 enterprise RL environments listed: one available now and 23 in development, according to the vendor.
24 listed itemsVenkat Rangapuram
AI research/data/engineering org plus a domain-expert network (per company)
Company size 500+Pack, price, samples & audit · unlock after NDA
Replicas of workplace apps with cross-app tasks and GUI/MCP interaction.
11 listed itemsNirmal Krishnan, Alexander Fung
Founders: Johns Hopkins computer science, ML and computational-genomics research
Company size 11–50Pack, price, samples & audit · unlock after NDA
Simulation Lab and tasks for software engineering, cybersecurity and enterprise work.
1 listed itemNazneen Rajani, Soumyadeep Bakshi
Hugging Face·Salesforce post-training research · Deloitte AI practice
Company size 11–50Pack, price, samples & audit · unlock after NDA
Computer-use environments with task and verification artifacts across desktop applications and cross-app workflows.
6 listed itemsFrancesco Bonacci, Dillon DuPont
Microsoft Windows Agent Arena research and engineering background
Company size 1–10Pack, price, samples & audit · unlock after NDA
Coding environments and long-horizon tasks, alongside the public DeepSWE benchmark.
2 listed itemsSerena Ge, Charley Lee
Waterloo CS · Cohere synthetic-data research, Google internship
Company size 11–50Pack, price, samples & audit · unlock after NDA
Privately built tasks compatible with Terminal-Bench and SWE-bench formats; independent of those benchmarks.
3 listed itemsShahzaib Awan, Shehryar Awan
Company size 51–200Pack, price, samples & audit · unlock after NDA
Packaged financial environments, tools and verifiers for long-horizon workflows.
2 listed itemsTudor Popescu, Jai Trivedi
Team with hands-on lending, finance-research and engineering backgrounds
Company size 1–10Pack, price, samples & audit · unlock after NDA
Open-ended interpretability research tasks for training AI research agents.
1 listed itemAnish Tondwalkar, Daniel Moon
Daniel Moon: MIT EE BS, EECS MEng
Company size 11–50Pack, price, samples & audit · unlock after NDA
Quantitative research environments where agents experiment, validate and improve.
1 listed itemRui Wang, Michael Zhang
ETH Zürich · G-Research·TransMarket Group·Etched
Company size 1–10Pack, price, samples & audit · unlock after NDA
Custom environments for browser-based enterprise work and terminal use.
2 listed itemsGuohao Li
Guohao Li: KAUST CS PhD, Oxford postdoc · research at Intel ISL, ETH Zurich, Kumo AI
Company size Not disclosedPack, price, samples & audit · unlock after NDA
Long-horizon software development environments with running services, databases and deployment tools; samples on request.
1 listed itemJoseph Wang, Sid Patllollu
Team of researchers and engineers (individual backgrounds not disclosed)
Company size 1–10Pack, price, samples & audit · unlock after NDA
Large off-the-shelf pre-/post-training and RL datasets plus custom petabyte-scale web-data procurement.
2 listed itemsAlana Xiang
All-technical team from Stanford, MIT and Magic.dev
Company size 1–10Pack, price, samples & audit · unlock after NDA
Suppliers that build or supply RL environments and tasks. Public benchmarks are separate from commercial offerings. Confirm scope, prices and sample access with each supplier.