Opptra - AI Product Engineer

About Opptra :

Opptra is building the worlds first AI-operated global distributor. Founded by Binny Bansal (Co-founder of Flipkart, scaled 0 $16B, acquired by Walmart), we own and execute all seven stages of retail operations across regions as the brand owner in each country.

The Role :

Opptra is an AI-first company helping international brands expand into Asia across India, GCC, and SEA through franchising, licensing, and inventory-led partnerships,(flip We move fast, build with intent, and run lean.

Were not hiring a PM and an engineer separately. Were looking for one person who can do both talk to a warehouse ops lead, understand the problem they're describing, sketch a solution, and ship it within the same sprint. You will gather requirements, design flows, write code, deploy, and iterate in direct partnership with business and functional teams.

- Sit with business and ops teams, run discovery conversations, and translate fuzzy pain points into clear product specs - without a PM in between.

- Design lightweight AI-native solutions : agents, RAG pipelines, automation workflows, or internal tools - whichever solves the problem best.

- Build full-stack : React/Next.js frontends, Python or Node backends, BigQuery-connected APIs, and LLM integrations using OpenAI, Anthropic, or open-source models.

- Deploy and maintain your own work - CI/CD, monitoring, iteration - not hand it off to someone else to ship.

- Iterate in the open : share weekly updates with stakeholders, incorporate feedback quickly, and kill features that dont work without ego.

- Participate in the AI roadmap - youll have a real say in what gets built next, and why.

- AI / LLM : You build with LLMs not just use them. You understand prompt design, model selection tradeoffs, RAG architecture, agent orchestration, and how to evaluate AI output quality.

- Frontend : React or Next.js, component design, data-dense UIs (dashboards, tables, filters) you build interfaces that non-technical users actually want to use.

- Backend : REST APIs (Python / Node), database design, BigQuery or similar, event-driven patterns, and integrations with third-party APIs (marketplace, logistics, catalog feeds).

- Deployment : Docker, basic CI/CD, cloud hosting you can get your own code to prod without infra support.

- Data literacy : You can query, understand, and act on data SQL fluency and a working instinct for metrics.

- Ships, doesn't just specify : Evidence of real products in production matters more than design docs or decks.

- Owns the full loop : If your AI system makes a bad decision at 2am, that's your decision. You monitor, you fix.

- Comfortable with discomfort : Scope isn't always clear. You figure it out rather than waiting for complete requirements.

- Learns in public : You share half-built things, ask for input, and improve you don't disappear for a month and emerge with something.

- A founding team that is hands-on including Binny.

- Real business problems to solve inventory, pricing, catalog, channel ops not synthetic exercises .

- AI-first tooling as default : Cursor, Claude Code, n8n, LangGraph you build the way good engineers build now.