Lead Product Data Analyst - AI Products

Responsibilities: Own instrumentation coverage for the AI pillar -define the events and properties each AI surface must emit, and drive them into the roadmap with PMs and engineering. Define feature-level KPIs and success metrics for Voice AI, Conversation AI, AI Employee, and Ask AI -what "working" and "adopted" mean, made explicit and trusted. Partner with the Experimentation and Causal Inference lead to design and read AI experiments, including measurement approaches for non-deterministic, fast-iterating systems. Separate real signal from instrumentation gaps, novelty effects, and data maturity in every read. Set the measurement direction and standards for the AI analytics domain -the canonical metrics and instrumentation contracts other teams build on -and mentor analysts as the pod grows. Translate findings into clear recommendations for AI PMs and influence the roadmap without owning it. Flag data gaps to Analytics Engineering and help shape the event taxonomy AI analytics depends on. Use AI tooling (Claude and similar) to move faster on exploration, documentation, and analysis.

Requirements: 7+ years in product analytics, data science, or applied statistics, with hands-on ownership of a product area's metrics end to end. Strong instrumentation instinct -you've defined event tracking/taxonomy and driven it into a product roadmap, not just consumed existing tables. Ability to define success metrics for a product from a standing start and get stakeholders to adopt them. Fluency partnering on experiments (A/B design, guardrails) and interpreting results honestly. Strong SQL and working proficiency in Python; comfort in a Snowflake + dbt environment. Comfort working amid imperfect, in-progress data -you consume governed sources and help raise the bar rather than rebuilding pipelines. Cross-functional influence -you shift PM priorities without direct authority in a fast-moving environment.

Nice to Have: Experience measuring AI/ML or LLM-based product features, including non-deterministic systems. B2B SaaS, CRM, or product-led growth background. Experience with conversational, voice, or agent/assistant products. Familiarity with Statsig or a comparable experimentation platform. Exposure to AI-assisted analytics workflows. Success in this role looks like: The AI surfaces are instrumented to a standard that makes them measurable, with coverage gaps closed on a known schedule. Each AI product has agreed, trusted feature-level KPIs and success metrics everyone uses. Adoption and quality of AI features are understood by segment, not guessed at. AI experiments are designed and read rigorously despite non-deterministic behaviour. AI PMs make roadmap calls off the analysis, and the pillar has a foundation the next analysts can build on.