Senior Product Manager

Job Description Job Profile: Senior Product Manager - Backend & Platform (CREW) Location: Whitefield,Bangalore, Karnataka- Onsite Experience: 6-9 years in product management, with at least 3 years owning backend, platform, or transactional systems. About CREW CREW is a personal travel concierge for the premium Indian travellers who want the trip to be great, and would rather not spend three weeks in browser tabs getting there. Two promises hold it up. We guard the wallet: member rates off directly contracted supply. And we stay with the customer: a real team of AI and humans from the first message to the flight home, handling the planning, booking, web check-in, the early check-in, the layover, the 6am flight change. Every travel platform in India stops at the booking. Nobody is playing for what comes after, which is the actual trip. That is also where CREW lives. Why this role matters - Landing the concierge feeling is a big backend problem. - Guarding the wallet is a sourcing and pricing system. Contracted supply, live rate shopping against public OTA prices, and the ability to prove the claim on every booking instead of asserting it in an ad. - Staying with customers is an orchestration problem. One trip is a visa, flights, a stay, transfers, activities and cabs, plus whatever the customer already booked somewhere else. - The economics are decided here too. How do we build systems with the perfect balance between AI and humans? What will you get to do here? - Sourcing, pricing and supply : The systems that make the price promise structurally true. Contracted and aggregator supply across hotels and flights, rate and availability freshness, price benchmarking that keeps the promise honest, failover when a supplier drops, and catalog quality. - Trip orchestration and the transaction layer :A trip is a dozen bookings that have to behave like one thing. Own the model that holds a trip end to end, keeps every piece in sync as plans change, and surfaces what still needs booking. That includes the money: confirmation, amendment, cancellation, refund and vendor payout, with reconciliation that proves every rupee is accounted for. This is a greenfield problem and it is yours to shape. - The AI execution layer :CREW's agents gather requirements, plan, call vendors and complete bookings. You own what sits underneath: the tools agents can call, how those tools are versioned and evaluated, guardrails and fallbacks, cost and latency per session, and the handoff to a human captain. The target is that most requests never need a person, and the ones that do reach one with the work already done. Cost per order is the scoreboard. - ​​​​​​​ Reliability and instrumentation :Define what healthy means for every backend surface: booking success rate, supplier error rates, response and resolution time, automation rate, cost per order. Build the event model that makes them measurable, then run the loop that improves them. Speed and handoffs are where customers judge us hardest, and both are a systems fix rather than a staffing fix. What we're looking for - 6-9 years of product management experience, with at least 3 years on backend, platform, or transactional systems. - Business instinct - You can understand the details of how a business works, makes money and where it leaks, and you can look at a supplier contract, a failed booking or one more automation and say what it does to margin. Platform work that never shows up in revenue, cost or retention is not work you enjoy doing. - Technical depth - You can hold your own in architecture reviews, API design, data modelling and failure-mode discussions. You do not need to write the code, but you should know what a bad state machine costs six months later. - Transactional systems experience - You have worked on something where a failure meant lost money, a broken order, or a customer stuck mid-transaction. Travel, commerce, payments, logistics or fintech are the closest fits. - Partner and integration fluency - You have built on third-party APIs where the partner is unreliable, the contract is rigid, and the customer still has to get a clean experience. - Systems mindset - You design reusable platforms, not one-off features, and you can tell the difference when a deadline is pushing you the wrong way. - Analytical rigour - You define the metric, check whether the data supports the claim, and prioritise on evidence rather than volume of requests. - AI-proven or seriously AI-curious - Prior work on LLM-powered products (agents, tool calling, evals, prompt and cost optimisation) is a strong plus. If not, show real depth of understanding and a track record of learning fast. - Comfort with ambiguity - Most of this is not built yet. You will write the spec, not receive it.