The role

You own how AI models actually run in production. Not building the models, running them well. You know how models get integrated at the API level, how to keep them up, quick and inside budget, and the best practices for operating them at scale.

You have done this before: you have kept model-backed features live, watched them under load, and fixed them when they slowed down, got expensive or started failing.

What you will do

Integrate and operate AI models at the API level: hosted provider APIs and self-hosted endpoints.

Own latency, throughput, cost and uptime for model-backed features.

Set up retries, timeouts, fallbacks, rate-limit handling and caching so features stay reliable when a provider degrades.

Manage API keys, quotas, model versions and rollout of new model versions without breaking things.





Track cost per request and per feature, and find the cheaper path that keeps quality acceptable.

Write the runbooks and best practices the rest of the team follows.

What we are looking for

One to two years running model-backed or API-heavy features in production, ideally at a GenAI product company.

Solid grasp of how model APIs work: tokens, context windows, streaming, rate limits, batching.

Comfortable in Python or Node, with basic backend and cloud skills.

You have handled a production incident on an AI or API feature and know what caused it.

You think in trade-offs: speed vs cost vs quality.

📌 AI Model Operations Engineer (Noida)
🏢 Hungama
📍 Noida