Agents You Can Trust : A COLM Happy Hour
Hear from AI leaders at Google, Meta, Salesforce, Glean, and more on what it actually takes to deploy AI agents safely in production.
Join Vinay Rao, VP at Google, for a fireside chat with Anish Das Sarma, CEO of Reinforce Labs, followed by a panel featuring:
Nilou Salehi | Professor, UC Berkeley & CEO, Across AI
Panagiotis Papadimitriou | Trust & Safety, Meta
Nilesh Dalvi | Engineering Leader, Glean
Manjeet Singh | Sr. Director, Agentforce AI, Salesforce
Moderated by Prasanna Desikan | Head of Healthcare AI, Centific
AI agents can look impressive in demos. Deploying them into real products, where they interact with users, data, and tools, is a much harder problem.
As agent capabilities increase, teams are grappling with a new set of questions: How do you evaluate behavior before deployment? What guardrails actually work in production? Where should accountability sit when an agent takes an unexpected action? And what does it take to earn user trust?
During COLM week, Centific and Reinforce Labs are bringing together researchers, builders, and product leaders for a candid conversation about the practical challenges of deploying agentic AI responsibly.
We’ll discuss agent evaluations, red teaming, guardrails, production safety, accountability, and user trust, with perspectives from teams building and deploying AI at scale.
Come for the speakers. Stay for the conversation.
Schedule
6:00–6:30 PM | Registration & refreshments
6:30–6:45 PM | Welcome & introductions
6:45–7:30 PM | Fireside chat with Vinay Rao, VP, Google
7:30–8:15 PM | Panel: Deploying Agentic AI Responsibly
8:15–9:00 PM | Networking
📍 San Francisco (exact address will be shared after you register)
🗓 October 8, 2026
Attendance is capped at 50 guests. RSVP early to secure a spot.
About Centific
Centific is a global AI data and services company backed by more than two million subject matter experts, providing data infrastructure for some of the world’s most capable AI systems.
Its research arm, Centific AI Research (CAIR) Labs, works across AI agents, reasoning, post-training, safety, evaluation, and alignment, including building and evaluating RL environments for enterprise AI agents. CAIR regularly publishes at leading research venues including ICML, NeurIPS, COLM, KDD, and ICLR.
About Reinforce Labs
Reinforce Labs is a Palo Alto AI startup helping enterprises find how AI agents fail, fix them at the source, and guard them in production. Chatbots, copilots, single and multi-agent systems: from a model behind an API to an agent in front of real patients, regulators and executives.
The team’s research has also appeared at leading venues including ICML and KDD.