Product Manager — Agentic Analytics

We're looking for an AI-Native Product Manager to own one or more of our core agentic analytics product areas. This is not a traditional PM role where you write specs and hand them off. This is a role where you'll be deeply in the product, using AI tools daily, making autonomous calls, and shipping continuously.

You'll own the full lifecycle — from understanding why a pharma brand manager can't trust an AI insight, to defining the product, to shipping it, to measuring whether it actually changed a decision.

You'll work across one or more of Tellius's core product pillars:

• Conversational Analytics — Natural language interaction with enterprise data. Users ask questions; AI agents investigate autonomously and deliver insights with reasoning and evidence. Handles both structured data warehouses and unstructured documents.
• AI-Generated Interfaces — Moving beyond static dashboards to dynamic, context-aware visualizations. The system generates the right interface for each specific insight — whether that's summary cards, interactive charts, or narrative explanations.
• Onboarding & Setup Automation — Reducing time-to-first-insight from days to minutes. AI-powered workflows that understand data relationships, suggest transformations, and validate configurations before users go live.
• Cross-Platform Intelligence — Bringing Tellius's AI capabilities into the tools users already work in. Ambient assistance that understands context across different enterprise applications.
• Vertical Intelligence — Pre-configured domain knowledge for specific industries (pharma, CPG, finance) — understanding their metrics, workflows, and decision-making frameworks out of the box.

You won't own all of these simultaneously. You'll own the ones that match your experience and where you can drive the most impact. But you'll need to understand how they all connect — because Tellius is a platform, and the product decisions in one area affect all the others.

This isn't a role where "AI" is a buzzword in the job title. We mean it operationally:

You will use AI tools to do your job. Prototyping, research synthesis, competitive analysis, spec drafting, data exploration — all of it. If you're not reaching for an AI tool first, you're working the wrong way here.
You will build AI products. The products you ship involve autonomous agents making decisions, surfacing insights, and taking actions. You need to understand — at a working level — how LLMs reason, where they fail, how to design for uncertainty, and how to build trust into AI-driven workflows.
You will think in systems. Agentic analytics isn't a single feature. It's a web of interconnected capabilities: memory, semantic layers, reasoning chains, action workflows, and governance. Your job is to make these systems work together coherently for real enterprise users.

1.  Deep AI Product Experience: You've shipped products where AI/ML is the core interaction, not a bolt-on. You understand the difference between "AI-powered" and "AI-native." You've wrestled with hallucination, confidence calibration, and explainability in production — not just in theory.
2.  AI Evaluation & Testing Expertise: You've built evaluation frameworks to test and validate agent behavior at scale. You understand how to use LLM-as-a-judge for automated testing, create human evaluation rubrics, design test scenarios that catch edge cases, and measure output quality beyond simple accuracy metrics. You know the difference between vibes-based testing and rigorous validation.
3.  Enterprise Analytics Domain Knowledge: You know what a pharma commercial analytics director actually does on a Tuesday. You understand semantic layers, data governance, metric definitions, and why an enterprise user needs to audit an AI's reasoning before acting on it. You don't need to write SQL, but you need to understand why the query matters.
4.  Agentic Systems Thinking: You can design product experiences where AI agents work autonomously — and where humans need to understand, guide, and override that autonomy. You think about agent memory, multi-step workflows, proactive vs. reactive triggers, and how to show reasoning without overwhelming users.
5. Speed and Autonomy Tellius is a 75-person company moving fast. You'll make calls without perfect information. You'll ship before it's polished. You'll iterate based on real user feedback, not months of research. You need to be comfortable with this pace and thrive in it.
6. Cross-Functional Collaboration You'll work closely with 44 engineers, 3 other PMs, designers, and a small GTM team. You need to be the person who connects engineering capability to business needs. Sales and CS are your frontline intelligence — you should be in their world regularly, not just at QBRs.

• 5+ years of product management experience, with a majority in growth-focused roles
• You've built products in analytics, BI, data platforms, or AI/ML
• You've worked in a company serving regulated industries (pharma, finance, healthcare)
• You understand and can articulate the difference between a dashboard product and an agentic product
• You've designed for AI transparency and explainability
• You can prototype and use AI tools fluently in your workflow
• Technical background or ability to work effectively with engineering teams
• You ask, "What decision does this enable?" before "What feature should we build?"

• A PM who treats AI as a feature checkbox
• Someone who needs 6 months of onboarding before shipping anything
• A process-heavy PM who needs a fully defined roadmap before starting work
• Someone who can't navigate ambiguity — our product surface is expanding rapidly

Understand the user deeply. Talk to pharma commercial managers, CPG RevOps teams, and finance analysts. Understand their workflows, their trust requirements, and their relationship with data. Don't rely on secondhand intel — be in the room (or on the call).
Define and ship product. Write lightweight specs (not 20-page PRDs). Work with engineers to understand technical constraints and possibilities. Make trade-off calls. Ship. Measure. Iterate.
Use AI to move faster—prototype with AI tools. Synthesize research with AI. Draft specs with AI. Explore data with AI. If you're doing something manually that AI could do, stop and fix that first.
Think about the platform. Every feature you build connects to the broader Tellius platform. A decision about how Kaiya surfaces insights affects how Apps renders them, which affects how the Browser Extension presents them. Think in systems.
Drive outcomes, not features. Your success metric isn't "shipped 12 features this quarter." It's "influenced 50 enterprise decisions" or "reduced time-to-insight from 20 hours to 30 minutes." We measure value by what changes in the real world, not what ships in the product.
Own trust and governance. Enterprise users — especially those in the pharmaceutical and financial industries — need to trust AI outputs. You'll think about explainability, audit trails, data lineage, and confidence communication as core product requirements, not afterthoughts.
Culture: We're an AI-first, outcome-focused, and lean organization. We don't do unnecessary meetings. We do write things down. We move fast and iterate. We trust each other to make calls. And we care deeply about building products that enterprise users actually trust.

• Competitive base salary + equity (Series stage, early growth)
• Full benefits package

Learning budget for courses, conferences, and tools

We don't want a cover letter. We want to see how you think.

Send us:

1. A product you've shipped that involved AI as a core interaction — what was the hardest trust or explainability problem you solved?
                                             OR
2. If you had 30 days at Tellius, what's the first thing you'd investigate and why?

That's it. Keep it short. We'll know if you're the right fit.