Senior Product Manager,
AI & Agentic Systems

Role Summary

We are looking for a product leader who can navigate the shift from deterministic software to probabilistic AI systems. You will not just manage a backlog; you will define how our AI interacts with the world.

You will partner with research and engineering to translate state-of-the-art model capabilities (GenAI, LLMs, Agents) into reliable, scalable, and high-impact business solutions. You will own the product lifecycle from "vague research capability" to "production-grade deployment."

Key Responsibilities

1. Strategic Definition & "The Art of the Possible"

Define the "Why" and "How": Translate abstract business goals into concrete AI formulations. Determine when to use a simple heuristic, a fine-tuned model, or a complex agentic workflow.
Manage the "Maybe": Champion a product culture that embraces uncertainty. Design user experiences (UX) that gracefully handle model hallucinations, latency, and non-deterministic outputs.
Unit Economics of Intelligence: Model the cost-to-serve for inference at scale. Balance model performance (accuracy/recall) against latency and cost constraints.

2. Technical Execution & Evaluation

Own the "Golden Set": You are responsible for the definition of "quality." Build and maintain robust evaluation frameworks (evals) to quantitatively measure model improvements.
Data Flywheels: Design the product loops that turn user interaction into training data (RLHF, automated data curation).
Agentic Orchestration: Define the logic for AI agents—what tools they can access, how they maintain context, and how they recover from failure.

3. Cross-Functional Leadership

Bridge Research & Engineering: Act as the translator between data scientists and business stakeholders.
Responsible AI & Safety: Proactively map adversarial risks and bake safety guardrails into the product definition.

Elite Qualifications

Experience

6+ years in PM, 3+ years specifically shipping AI/ML products to production.

Technical Fluency

Understand trade-offs between RAG, Fine-Tuning, and Context Caching. Vector DBs, latency, and tokens.

Probabilistic Mindset

Ability to build products where the 'happy path' is not guaranteed. Mitigation of AI error.

Analytical Rigor

Proficiency with SQL, Python, Tableau. Pulling own logs to analyze model behavior.

Ready to Build?

We are only accepting applications from high-agency individuals who can demonstrate a track record of institutional-grade AI deployment.

Apply for Role

Interview Process

Institutional Alignment Call

Technical Intelligence Assessment

Architectural Deep Dive

Founder's Vision Session