Senior Backend Engineer — Verification Platform

• A verification pipeline that cross-references merchant marketing claims against specs reviews and transaction data. Latency throughput and accuracy are in tension - you'll navigate those tradeoffs daily.
• Agent infrastructure where AI agents consume verified knowledge via MCP and API. Each agent has different context budgets latency requirements and trust thresholds. The platform must serve them all.
• A cost-aware compute layer. Every LLM call has a dollar cost. You'll own the per-query cost ledger KV-cache hit rate optimization and the thin LLMAdapter interface that lets us swap providers without rewriting the stack.
• A system under rapid evolution. New categories new agent patterns new verification methods ship weekly. Your architecture must absorb change without accumulating debt.

• Verification infrastructure. The pipelines that ingest raw product data and transform it into verified knowledge. Cross-referencing marketing claims against ground truth at scale. You own correctness freshness and cost.
• Agent-serving platform. The API and MCP layer that delivers verified intelligence to AI agents and consumer surfaces. Schema design rate limiting context-window-aware response shaping. You'll build what agents trust.
• The LLMAdapter and cost architecture. A thin interface that pins model versions manages provider portability and instruments per-PR cost impact. Every PR you review includes a cost delta. You'll own the seven-component verification dashboard that makes this visible.

• How you think. You reason about systems in terms of invariants failure modes and tradeoffs - not features. When something breaks your first instinct is "what's the structural cause?" not "what's the quick fix." You write specs before code and tests before implementations. You think about what happens at 10x the current scale.
• How you work. You ship code that other engineers can read extend and trust. You review PRs for cost impact not just correctness. You build thin abstractions that earn their complexity. AI is part of your workflow - you use Claude Code or Cursor to accelerate but you verify what they produce. You've maintained systems in production where your decisions had real cost and reliability consequences.
• What you've probably built. Data pipelines API layers verification systems or ML-adjacent infrastructure at meaningful scale. You've operated what you built - not just shipped it and moved on. You can name the on-call incident that taught you the most. We care about the system and the reasoning not where you built it.
• Who this isn't for. This role is wrong if you mainly want to work on greenfield features without owning operations. It's wrong if "that's the infra team's problem" is part of your vocabulary. It's wrong if you prefer thick abstractions and framework-driven development over thin purpose-built systems. You'll thrive here if you like owning the full stack from schema to deployment to cost monitoring.

• Total first-year comp: $400000 - $500000 (base + equity + profit sharing).
• Base: $280000 - $340000. Top of market.
• Equity: Profits Interest Units (PIUs) - Class B Membership Interests at $0 strike price. Actual ownership from day one. Capital Gains tax treatment.
• Profit sharing: Annual pro-rata share of free cash flow. Real cash every year not a promise tied to an exit.
• Liquidity: Annual tender offer - turn ownership into cash every year. No waiting for an IPO.
• Benefits: 100% premium coverage for you and your family. Unlimited PTO that we actually use.

What We Do

Product.ai (formerly Demand.io) is the truth layer for commerce. Built on Axiomatic Intelligence — a proprietary adversarial reasoning methodology that stress-tests product claims against physics economics and engineering constraints — Product.ai delivers verified purchase verdicts not summaries. Product.ai tells consumers when NOT to buy. Product.ai emerges from Demand.io a profitable bootstrapped AI commerce company whose SimplyCodes platform processes over $1B in annual transaction value with a team of 20. Founded by Michael Quoc.

Gallery