Job Title: Senior Staff AI Engineer - Grade 13 [Senior Arch] About the Team: The AI Solutions Org is a new, high-velocity engineering unit at Flipkart with a singular mission: to accelerate AI adoption across the entire company. Will operate as a SWAT team for AI - building end-to-end solutions for business functions that lack engineering bandwidth, while also partnering with the core Data Science team to optimize and scale state-of-the-art models (FK LLMs, Open Source, Commercial).

About the role: A Senior Staff AI Engineer operates at a different scope from Staff. The problems you own are not well-defi ned — they are multi-disciplinary charters where DS, Engg, and Product have competing definitions of success, and where the goal itself changes as you learn. Your job is not to execute a plan; it is to defi ne what the right plan is, hold the technical direction through iteration, and coordinate across functions without deferring to any single team's framing.

You set the technical direction for initiatives, not just systems. You make the call when DS says one thing, Engg says another, and Product wants to ship in four weeks. You design the evaluation framework that the entire org uses.

You decide which techniques are worth the team's time to investigate. You are the person other senior engineers come to when the problem is hard enough that the usual playbooks don't apply. The bar for Senior Staff is not does excellent technical work.

It is: does this person change the technical trajectory of a team or initiative, and do they make the organisation smarter about AI by being in it What you will do:

- ● Own ambiguous, cross-functional AI charters — where the goal itself must be negotiated across DS, Engg, and Product,



and where the definition of done evolves across iterations. ● Set the technical direction for multi-team initiatives. Make architectural decisions that other teams build on. Hold that direction under pressure from shifting requirements, competing priorities, and partial information. ● Design evaluation strategy at the platform level: how does the org know if its AI systems are working Who owns each metric, how are disagreements resolved, and how do you detect silent degradation at scale ● Coordinate across DS, Engg, and Product without deferring to any single team's framing. Recognise when three teams are describing the same root cause differently and design a resolution process rather than picking a side. ● Set the technical bar for the org: through design review, evaluation standards, architecture patterns, and the questions you ask that others don't. ● Make defensible decisions under uncertainty. When the data is incomplete and the timeline is short, you make a call with explicit conditions, design a fast-path investigation, and own the outcome. ● Shape what the team investigates. Filter the AI research landscape. Decide which techniques are worth prototyping, which are not, and why — and communicate that judgment clearly enough that others can apply the same reasoning. ● Identify where AI creates genuine leverage vs. where it is the wrong solution.



Push back on AI-fi first thinking when the problem calls for something simpler or different.

What we are looking for :

- Ownership at charter scale ● You have owned a cross-functional AI charter — not just a system. The goal itself changed over iterations. You drove the definition of success, not just the execution. ● You have experienced a direction failure — not just a component failure. You pivoted the charter, owned the decision, and can explain the systemic cause. ●Design production-scale AI infrastructure: RAG pipelines, vector databases, LLM gateways, semantic caching, and real-time inference serving. ● Optimise models for production: quantisation, distillation, ONNX/TensorRT conversion, GPU utilisation, and cross-language serving (Java/C++).

Background we expect

- ● 14+ years of engineering experience, with at least 3-5 years working primarily on production AI/ML systems. ● Demonstrable ownership of at least one cross-functional AI initiative where you drove the problem definition, not just the solution. ● A track record of technical decisions that other teams built on — architecture patterns, evaluation frameworks, or platform choices that outlasted your direct involvement. ● Experience coordinating across DS, Engg, and Product on a high-stakes AI initiative — resolving competing priorities, not just facilitating alignment. ● Deep fluency in the AI stack: from training and fine-tuning decisions through serving, evaluation, and production monitoring. ● Published work, open-source contributions, or internal technical leadership that is visible and cited beyond your immediate team is a solid positive signal — not a requirement.

📌 Senior Architect - AI (Bengaluru)
🏢 Flipkart Marketplace
📍 Bengaluru