Trillium Labs

SkyRL is an open-source reinforcement-learning training framework for AI models. Nathan Lambert said Trillium plans to use it as part of its open post-training stack.

rLLM is an open-source framework from UC Berkeley's Sky Computing Lab for training language agents with reinforcement learning. It connects agent code to evaluation and training through a decorator-based Python API or CLI workflow: agents run in sandboxes, the model gateway captures token IDs and log probabilities, traces are organized into episodes and trajectories, evaluators compute rewards, and a training backend updates the model. It supports existing or custom agent harnesses, Docker, Daytona, Modal, and local sandboxes, with training backends including verl, tinker, and Fireworks, and methods including GRPO, REINFORCE, RLOO, supervised fine-tuning, and on-policy distillation. The project also provides integrated benchmarks and requires Python 3.11 or later.

Snorkel AI is an AI data-development company that builds specialized training datasets, benchmarks, and evaluation environments for frontier models and agents, particularly in high-stakes and specialized domains. Its work addresses distributional gaps, benchmark blind spots, and tasks where correctness is difficult to define. The company also publishes research and operates Open Benchmarks Grants; the video specifically describes its collaboration with UC Berkeley on FinQA, an expert-validated financial question-answering dataset built from SEC 10-K filings.

Braintrust is an AI observability and evaluation platform for agents and other AI systems. It traces production prompts, responses, tool calls, latency, cost, and quality; supports searching logs, real-time trace inspection, dashboards, and custom views; and runs experiments against versioned datasets with scoring by LLMs, code, or humans. Teams can turn production traces into evaluation datasets, compare prompts and models, investigate recurring failures through natural-language queries, and use the results to identify regressions and improve agents. Braintrust provides native SDKs for Python, TypeScript, Go, Ruby, C#, and other environments, an MCP server for querying logs and running evaluations from an IDE, and Brainstore, its database and query engine for AI data. The platform states that it supports framework-agnostic integrations, SOC 2 Type II certification, GDPR and HIPAA compliance, SSO, role-based access control, and hybrid deployment options.

Ramp Labs is Ramp's applied research group and research blog, covering AI systems for finance and related agent technologies. Its published work includes post-training Qwen 3.5 models for financial question answering, KV-cache compression and cartridge techniques, large agentic swarms for penetration testing, transfer between post-trained models, self-improving monitoring, and semantic tagging. Ramp Labs also publishes evaluations such as Ramp Accounting Bench, which tests models on accounting tasks using professionally created workflows and grading rubrics.