SDE2 - API Platform

present of AI; Bengaluru; Presencial; INR 900.000 - 1.500.000; Jornada completa

Descripción de la vacante

Sarvam AI seeks a senior backend engineer to own core API platform components that serve language, speech, vision, and translation models for developers. You will work with Python, asyncio, and streaming systems while practicing reliable, scalable design on Kubernetes.

The role emphasizes latency, testing under real load, and a practical approach to metering and usage-based billing for AI services. Bengaluru-based, on-site role with growth in a fast-evolving AI platform.

Formación

• Production backend experience with Python
• Willing to learn Go or pick up new tech
• Experience with streaming, caching and databases; Kubernetes ops and reliability testing

Responsabilidades

• Own core parts of an API platform serving language, speech, vision and translation models
• Support streaming responses and latency metrics at scale
• Design for failure, debug live traffic and ensure protocol fidelity under load
• Work with metering and usage-based billing considerations for AI products

Conocimientos

Python

Asyncio

Streaming systems

Kubernetes

Herramientas

FastAPI

PostgreSQL

Redis

Descripción del empleo

The role is about owning core parts of an API platform that serves language, speech, vision and translation models to developers. A strong candidate has several years of production backend work in Python, some Go or a real appetite to learn it, and hands on experience with streaming, caching, databases and running services on Kubernetes. Reliability, performance and careful testing under real load matter more than titles.

How it fits the AI era

The work sits directly on top of AI models, so the platform is where model output becomes a usable product for other builders. AI tools can speed up coding and testing, but the judgment about routing, failover, backpressure and protocol fidelity stays with the engineer. The human part is designing for failure and debugging live traffic when things go wrong.

How it builds AI experience

This role builds practical experience in serving models at scale, including streaming responses, latency metrics and compatibility with common inference servers. It also gives exposure to metering and usage based billing for AI products, which is a growing area. Someone wanting this job could strengthen their case by building a small proxy or gateway in front of an open model and measuring its behavior under load.

Worth showing

• Python asyncio and FastAPI
• SSE and streaming systems
• PostgreSQL and Redis
• Kubernetes production operations

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Sarvam AI

Sarvam AI builds a full-stack sovereign AI platform for Indian languages, developed, deployed, and governed entirely in India (per the company).