Arize x Turing AI Builders Meetup: Optimizing Agents with RL Gyms and Prompt Learning
Join Arize and Turing for an evening exploring how to optimize agents with RL Gyms and Prompt Learning.
What to Expect:
Arize Talk: The Prompt Learning Loop
Prompt Learning, an Arize research-backed method for improving LLM behavior, introduces a lightweight, interpretable way to fine-tune behavior at inference time. It enables faster iteration, clearer accountability, and more flexible control—all with just English. Priyan Jindal, AI Engineer at Arize, will break down the motivation behind the approach, research benchmarks, and how it compares to traditional prompt engineering and fine-tuning.
We'll demo how English feedback can dynamically shape system prompts to improve agent accuracy, and walk through real examples where Prompt Learning outperforms scalar reward optimizers and enhances generalization—even in zero-shot settings.
Turing Talk: Reinforcement Learning (RL) Gyms
RL Gyms are transforming how we train AI agents shifting them from passive responders into dynamic, decision-making problem-solvers. These environments provide the infrastructure for AI to learn through interaction and feedback, simulating real-world complexity in a controlled, measurable way. Ankit Jasuja, Engineering Director at Turing, will explore the technical and strategic value of RL gyms: how they enable learning beyond static data, build long-term capabilities, and offer critical safety and alignment checks before deployment. We’ll also look ahead to the next frontier multi-agent, high-stakes simulations and how RL gyms are shaping the feedback-driven future of AI
Community Demos: If you are interested in sharing your work, please reach out to dcouzon@arize.com for availability.
Who should attend:
Whether you’re building agents, evaluating LLM performance, or optimizing real-world deployments, this session will offer a fresh lens on how language models can actually learn from language.
Agenda:
5:30pm - 6:00pm: Guest check-in & networking + food/bev
6:00pm - 6:20pm: Arize Talk
6:20pm - 6:40pm: Turing Talk
6:40pm - 7:00pm: Community Demos
7:00pm - 8:00pm: Networking
8:30pm: Event End