AI for mental healthcare

Build AI That Changes How Mental Healthcare Is Delivered.

Amaha brings together therapy, psychiatry, self-care, community, specialised programmes and hospital care. We're building AI across this ecosystem, working closely with our clinical teams to make care more accessible, personalised and effective.

We hire for what you can build, not for what your resume says.

The role

We believe technology should give clinicians more time to do what only they can do: care for people.

Your role will be to identify the administrative work that takes them away from care, and build AI systems that give them that time back. From documentation and follow-ups to everything in between, you'll build reliable systems that make care more efficient, consistent and scalable, securely and responsibly.

The clinical work

People, Context and Clinical Judgement

  • Understand a person's experiences and context
  • Make the judgment calls only a clinician can
  • Hold the relationship that makes care work

Where AI can empower

Documentation, Coordination and Administration

  • Turning sessions into clean, structured notes
  • The repetitive ops, support, and follow-up work
  • Surfacing the right thing at the right moment

What you'll build

Prototypes, automations, and agents that actually ship.

AI for Care

Tools that make providing and seeking care smoother, more personal, and more accessible.

RAGagentsMCPtool-calling

AI for Operations

Agents and workflows that take repetitive, multi-step work off our clinical, operations, and support teams, so more of their time goes to people.

workflowsevalsdeploymentintegrations

Who we're looking for

We're looking for engineers who enjoy building.

You might have built professionally, through side projects, hackathons, or your own experiments. We care more about how you approach problems and what you've built than your years or experience.

  1. You've built with LLMs and enjoy working with the systems around them, including RAG, agents, tool-calling, automations, and evaluation.
  2. You can take a problem from idea to deployment comfortable working across the stack with Python and/or JavaScript, APIs, and cloud services.
  3. You understand what it takes to build systems people can trust privacy, security, evaluation, and safe-by-default design are part of how you approach the work.
  4. You use AI in your own work and stay curious about how these systems behave. You naturally use it to write, experiment, review, and debug.
  5. You have things you've built and can show us how you think through repos, prototypes, and live links we can browse.

Here's how you can get started

Start With a Problem. Build Something Useful. Show Us How You Think.

No resume-first screening. We'd first like to see how you turn an idea into a working prototype.

Pick one, build it

  • A support assistant a conversational agent that helps someone figure out what kind of help they might need.
  • A Notes Copilot turn a sample session transcript into a clean summary with follow-up actions.
  • An automation take any repetitive, multi-step workflow and make an agent do it end to end.
  • Your own idea see a sharper problem you can tap into to build a solution? Build that instead.

Send us

  • A deployed link we can use
  • The GitHub repo
  • A 2–3 min Loom explaining what you built, the choices you made, and where you'd take it next
Data

Use synthetic, sample, or public data that you create or source yourself. This exercise doesn't require you to use Amaha data or real user information.

An overview of our process

Five steps. We move fast.

  1. You share your work

    Fill in the form and send us your build, deployed link, repo, and a Loom.

  2. We review it

    We'll look at all your submissions manually. No black box.

  3. Round 1 — you walk us through your build

    We'll talk through your architecture, trade-offs you made, and what you'd do differently.

  4. Round 2 — live practical exercise

    We'll work together on a problem (idea to deployment) similar to the kind of work you'd be doing here.

  5. Round 3 — fit check

    A conversation about how you work, what drives you, and how you think about doing this responsibly.

Build Something That Could Change How 200Mn Indians Access Mental Healthcare.

There's a lot to explore: care, clinical workflows, personalisation, agents, evaluation, safety and the systems that connect them. Start with one problem and see where you can take it.

Start your application

Have a question about the role or the process? Reach out at support@amahahealth.com or +91 20 7117 1501.