Uplift AI

A YC-backed Startup. We provide the best voice models for the 60% of the world that speaks regional languages, onboarding the next 100 million users to digital services.

upliftai.org

Forward Deployed Engineer

About the role

The voice agents of Uplift AI (YC S25) are starting to answer and make calls for Pakistani businesses, in Urdu, at the scale of a full call centre. Every company is different: its systems, its customers, the way those customers actually talk. You'll be the engineer who makes it work for each one.

That means walking into a real business problem with no playbook, figuring out what's actually going on, and building whatever solves it, from the integration to the internal tool to the fix in production. You'll work directly with the founding team of a Y Combinator-backed company whose models are already used by Khan Academy and more than a thousand developers, and you'll have the room to try your own ideas, test them, and rethink them when they don't work.

What you'll do

  • Get customers live. Connect Uplift's voice agents to each customer's systems, from order databases to CRMs, and make the whole thing work end to end.
  • Solve problems nobody has solved before. When a deployment breaks in a way no one expected, work out why from first principles and fix it.
  • Listen to real calls. Find where the agent struggles, figure out the cause, and make it better.
  • Build what's missing. If a tool, script or feature would make deployments faster, build it without waiting to be asked.
  • Be the link between customers and the core team. Turn what customers need into clear product changes.
  • Own your work in production. Monitor it, improve it, and stand behind it.

What you'll bring

  • You've built and shipped real things, at work, in a startup, in a hackathon or on your own, and you can walk us through the hard decisions.
  • You're a strong problem solver. You break problems down from scratch instead of reaching for the nearest tutorial.
  • You're comfortable across the stack: backend and APIs, databases, and enough frontend to build a tool someone can use.
  • You can talk to customers. You can explain a technical problem to a non-technical person, and listen well enough to find the real one.
  • You're comfortable in Urdu, because you'll be working with real calls in it.
  • You're driven and energetic. You set your own goals, move fast, and do good work without being prompted.

Nice to have

  • You're a recent graduate, someone with 1-3 years of experience, or still at university.
  • Experience with integrations and APIs from the other side: webhooks, third-party systems, messy data.
  • Comfort on the command line, with servers and with containers.
  • Any exposure to real-time voice, telephony or AI agents.

What you don't need

  • Seniority. We're looking at 0 to 3 years of experience.
  • Experience in voice AI. You'll learn the domain here. What we care about is how you think and how fast you learn.
End of description

What they're looking for

Must have

  • Python
  • REST APIs
  • PostgreSQL
  • React

Nice to have

  • Linux and Shell
  • WebRTC
  • Docker

Skills named in this posting

  • python
  • rest api
  • postgresql
  • react
  • linux
  • shell
  • webrtc
  • docker
  • crms
  • webhooks
  • containers
  • telephony
  • ai agents
  • urdu

About Uplift AI

  • San Francisco, CA
  • 1-10 people
  • upliftai.org

Uplift AI is a YC-backed startup that builds voice AI for Pakistan's languages. Its model, Orator, speaks Urdu with human-like realism, and Khan Academy has used it to dub 2,500 educational videos into Urdu. More than a thousand developers build on Uplift's APIs, and the team is now bringing AI calling to Pakistani businesses: voice agents that talk to customers in Urdu, at the scale of a full call centre. Uplift does the whole stack in-house, from its own data and labelling tools to its own models and voice infrastructure. It is backed by Y Combinator and Indus Valley Capital, and was founded by engineers who built voice and AI systems at Apple (Siri), Amazon (Alexa) and AWS.