Skip to main content

Careers

Help build the routing layer for AI. We are an early team solving hard problems in model routing and inference.

Why Inferbase

The work sits at the layer most AI applications depend on.

Inferbase classifies each request and routes it to the model that fits best on quality, cost, and latency, then serves it through one OpenAI-compatible API. No single model is right for every request, and the best pick changes from one task to the next.

We are early, so the systems you build and the bar you set will shape how the product scales. That means broad ownership and real influence over what gets built, alongside the ambiguity and shifting priorities that come with an early-stage team.

What we offer

  • Meaningful equity and real ownership of what you build.
  • A direct hand in a product at the core of AI infrastructure.
  • A small, senior team that moves quickly.
  • Substantial engineering and go-to-market problems, not narrow ones.
  • A team that works together in person, onsite in Bangalore.

How we work

A few things that are true of working here.

Own it end to end

You take systems from design through production and on-call. The code and how it behaves under real traffic are both yours.

Small and senior

A handful of experienced people who move quickly, with less process, more judgment, and a direct line to the founders.

Evidence over adjectives

We quantify a problem before building for it, and back decisions with data rather than opinion.

Correctness on the money path

Where money and data are on the line, we hold a high bar through review, testing, and instrumentation.

Onsite in Bangalore

We work in the same room on the same problems, in person and full-time.

How we hire

A short, respectful process, usually two to three weeks end to end.

  1. 01An introductory conversation.
  2. 02A deeper discussion of the role with the founders.
  3. 03A practical assignment or working session.
  4. 04A final conversation with the founders.
  5. 05An offer, usually within two to three weeks of the first call.

Start building with smart routing.

Automatically route workloads to the right model for every task, every time.