Lead Product Analyst - monday Agents

Description:

At monday.com, we help teams get more work done. We are the best AI work platform that empowers teams to automate, build, and scale their impact end-to-end with tools that actually execute the work for you. With over $1B in ARR, 250,000+ customers, and a global team, we're serious about building a product people love to use and giving our employees the same ownership and flexibility to shape the way the world works.

As Lead Product Analyst for monday Agents, you will own the analytical direction of the roadmap for our next-generation autonomous workforce ecosystem. You will be the primary analytical focal point for the group, turning complex, unstructured system data into strategic product recommendations. You will champion a user-centered analytics practice. You will build trusted relationships around monday.com Agents data across the organization, and help monday.com carry forward the learnings and ways of working that define how we operate in the agentic era.

Skills that will make you successful:

• The agentic mindset - Convert ambiguous, non-deterministic system behaviors into structured metrics; know when to use classic statistics vs. custom behavioral heuristics
• Master storyteller - Bridge deeply technical data science concepts and high-level product strategy; influence diverse cross-functional stakeholders
• User-centric rigor - Look beyond token counts; measure whether an agent actually solved a human problem efficiently

Lead the vision and strategy

• Build and own monday Agents' measurement foundation—what we instrument, how we evaluate outcomes, and the success metrics product and leadership use to steer the roadmap across the monday ecosystem.
• Partner with product leadership to turn abstract AI capabilities into structured, data-driven product roadmaps that prioritize user value and system reliability.

Drive innovation

• Design and run rigorous agent evaluation at scale - from "LLM as judge" setups to behavioral profiling - so we can benchmark what works, what fails, and where to invest next.
• Uncover deep behavioral insights on how humans interact with autonomous workflows, establishing new industry benchmarks for human–agent collaboration metrics.

Deliver and scale impact

• Lead cross-functional alignment across engineering, data science, and product to optimize agent accuracy, latency, execution success rates, and token consumption tradeoffs.
• Translate complex execution logs and multi-step agent trajectories into actionable, executive-ready insights that accelerate feature deployment.

Define ecosystem governance & trust

• Build the analytical frameworks needed to monitor agent guardrails, permissions boundary compliance, and security patterns, ensuring autonomous execution remains completely safe and enterprise-ready.
• Measure where custom agents drift from user intent, and turn those signals into a closed loop that continuously sharpens the AI Agent Builder.

• Lead-level product analytics - 6+ years leading analytics in product—ideally on agents, automation, or similarly complex surfaces where user outcomes matter more than funnel volume
• Leadership & ownership - Strong leadership and extreme ownership—you set direction, build trust across product and engineering, and drive clarity and outcomes end-to-end
• AI early adopter - You pick up new AI tools quickly, put them to work in your own analysis, and help the team unlock new capabilities as the stack evolves
• AI product evaluation - Practical experience defining how AI features are judged: outcome quality, reliability, guardrails, drift from intent, and the tradeoffs between cost, latency, and user value
• Technical foundation - Strong SQL and Python to explore agent execution data at scale; comfortable partnering with data engineering on instrumentation, not only consuming pre-built reports
• Trace-native measurement - You've turned multi-step execution data (runs, traces, JSON logs) into metrics, dashboards, and definitions that engineering and product align on and ship against—not one-off analyses