SESSIONS ON-DEMAND
AI Analytics for Builders
EXPERIMENTATION
Build an AI Growth Optimization System
What you'll learn
- Identify where evals end and experiments begin
- Layer evals, experiments, and analytics
- Automate your experimentation loop
Why this topic matters
Evals tell you your model is working. They don't tell you it's working for your users. The jump from offline evals to live experiments is where most AI product teams stall, and where the teams that don't pull ahead. This session shows you how to build the full loop.
Presenter
Eric Metelka, Director of Product Management at Statsig
EVALS
Agent Evals 101: Best Practices for AI Product Teams
What you'll learn
- Why agent evals can't wait
- A starting eval framework for AI product teams
- A 3-6 month maturity path for your eval practice
Why this topic matters
Agent evals are the difference between AI features that ship with confidence and ones that quietly degrade user experience at scale. Teams that build this muscle early will move faster, catch failures before users do, and have the data to make the case for what to build next.
Presenters
- Vinay Goel, Staff AI Engineer at Amplitude
- Ken Kutyn, Solutions Engineer at Amplitude with 10+ years in experimentation & AI evals.
AI PRODUCT TEAMS
North Star Metrics for AI Agents
What you'll learn
- How to stop tracking workflows, start measuring work.
- How to build & monitor Agent Success Rate.
- An eval rubric that anchors your release roadmap.
Why this topic matters
Every agent team hits the same wall. A feature wows in demos, ships to mixed reactions in production. Product Review is tomorrow and you can't answer the obvious question: is the AI actually doing good work? Engagement metrics don't help. AI doesn't produce a single happy path, it produces a wide distribution of outcomes. Discover the metric system every PM team needs before shipping an agent.
Presenters
- Sandhya Hegde, Co-Founder, Calibri Labs, Ex-Amplitude
- Justin Bauer, Co-Founder, Calibri Labs, Ex-Amplitude
AI PRODUCT TEAMS
Automate Your PM Workflow with Claude Code and Analytics
What you'll learn
- How to go from prompts to programs with Claude Code
- The top automation patterns you need for your PM workflow
- The PM quality checklist before you ship
Why this topic matters
Prompts are just the start. Learn how to use Claude Code to build automations that actually run—not just generate text. Plus: how to instrument your AI workflows so you can prove they're working. Go from "great prompt" to "this runs every week without me."
Presenter
Frank Lee, AI Product Leader at Amplitude, Founder at Inari
TEAM EVOLUTION
How to Stay Relevant as Product, Growth, and Marketing Blur
What you'll learn
- Expert takes on the product/growth/marketing overlap
- How to identify the skills that compound
- How to recognize new expectations for the job
Why this topic matters
The AI era isn't just changing individual tools, it's collapsing the walls between product, growth, and marketing. The practitioners who are staying ahead are the ones who are redefining what their job actually is. Join for practical insights on what the new job expectations are for modern teams, where the walls are coming down between different teams, and what skills compound in this new model.
Presenters
- Krista Seiden, Digital Analytics Leader, Product Strategist, and Industry Evangelist
- Adam Fishman, Advisor, Interim VP of Product at Mozilla, ex-Patreon, ex-Imperfect Foods
AI WORKFLOWS
Build AI Workflows That Actually Save PM Time
What you'll learn
- The Automate/Amplify/Skip Framework
- 3 Workflows You Can Run This Week
- Know When to Skip AI Entirely
Why this topic matters
Most PMs are using AI, but reaching for it reflexively, and spend more time crafting prompts than they save. This session will give you a decision framework and three battle-tested workflows so AI actually speeds up your week instead of adding more steps to it.
Presenter
Enzo Avigo, Amplitude, June.so, Intercom
AI ANALYTICS
AI Feature Analytics: A Practical Framework
What you'll learn
- Measure AI feature behavior
- Connect events to business questions
- Design a scalable AI event taxonomy
Why this topic matters
Most product teams are just starting to think about AI in their workflows—not yet building agents, but adding AI features and trying to figure out what to measure. This session is a practical deep dive into setting up the basics correctly. If your team is at the "we're adding AI features, now what do we measure?" stage, this is where to start.
Presenter
Glenn Vanderlinden, Co-founder, Humans37 | Data Strategy & Operations Expert
AI ANALYTICS
How to Stop Shipping Blind with AI-Built Products
What you'll learn
- Instrument your AI-built product in Amplitude
- Build a knowledge base that compounds
- Run a weekly learning rhythm
Why this topic matters
Most teams shipping AI products move fast and measure late. This session shows you how to close that gap: instrument your product from day one, build a shared knowledge base that compounds with every release, and run a weekly rhythm that keeps everyone connected to the same signals.
Presenter
Gregor Spielmann, Co-founder, Adasight | ex-Amplitude, Optimizely
AI ANALYTICS
Build an AI-Ready Analytics Program for Humans and Agents
What you'll learn
- The new behaviors you must measure
- Bot taxonomy and how to work with it
- A starter checklist you can act on
Why this topic matters
AI has rewired how people find products and use them, and a growing slice of your activity isn't people at all. Some bots improve outcomes. Others distort your data. And a growing share of your humans are working through Claude or ChatGPT while your product runs headless behind the scenes. Most analytics programs weren't built for any of this. This session covers what to pay attention to in 2026.
Presenters
- Jim Gordon, Award-winning analytics leader, also known as Jimalytics
- Jim Kultgen, Head of AI Strategy at Amplitude
PRODUCT VALIDATION
Validate Your Idea with Data Before You Build
What you'll learn
- Diagnose whether you're about to “vibewaste”
- Write a sharp, testable bet on your idea
- See the full path from bet to evidence-based verdict
Why this topic matters
Avoid spending weeks on the wrong idea by learning a 14-day validation framework that uses behavioral data to tell you whether your idea is worth building.
Presenter
Elena Avramenko, 15+ yrs in Product & Marketing, award-winning founder, 100+ new product launches
BUILDING WITH LOVABLE
How to Vibe Code with Lovable and Amplitude
What you'll learn
- Ship and instrument in the same flow.
- Where this stack actually wins.
- How to set it up once and not regret it later.
Why this topic matters
Most builders move fast in Lovable and add analytics later—which means two weeks of cleanup once real users show up. This session walks through how to ship a product you build in Lovable and instrument in the same flow, so Amplitude is capturing data from the first visitor and you're not backfilling event schemas after the fact.
Presenters
- Dmitrii (Dima) Abramov, Product Lead at Foundever, Lovable Ambassador, Vibe Coding Community Builder
- Jake Lee, Amplitude Solutions Engineer