AI Engineer (Contract)

With Saras, we're building the accountability systems of the future for busy professionals. We're starting out by solving standardized test prep in a way that creates real outcomes for real users, beyond the benchmarks, with ambitions to make a huge dent in this $120B+ market.

What we build through Saras feeds forward towards our even larger ambitions at Karwaan Labs: to bring frontier technology to an exciting space where it can produce meaningful outcomes beyond benchmarks. We're independent, bootstrapped, remote-first but mission-oriented.

The role involves dealing with a new class of software altogether, which is 80% configuration, 15% coding with agents and 5% tasteful intuition about systems. Which is why you'd be expected to learn quickly, make mistakes and have a fun time working on important problem statements that matter to real people doing real things. Hermes, Codex and the rest of the stack are your co-workers; be ready to build the way engineering will get done in the future.

About the role

• Own and evolve the platform that Saras runs on, custom-built with open source components. Make onboarding repeatable, and automate as much of the work as possible. By eliminating the recurring work over time, you create room for the higher-leverage parts of the charter
• Build messaging gateways that help Saras scale across platforms while preserving continuity. Saras lives on Telegram today, with WhatsApp and iMessage on the roadmap
• Build the systems around the agent: conversational intake, personalized study plans, routines, payments and integrations with proprietary 3P platforms
• Share your learnings through social channels and, where it makes sense, make open source contributions to high-profile projects. Build career capital and exemplify how to build outcome-driven agents

The stack is intentionally open-ended, but you'll work with autonomous AI agents, multi-tenant messaging, Docker, cron-driven automation, cloud deployment and agent frameworks in the Hermes/OpenClaw lineage. You'll also own the operational layer: per-profile logging, memory-integrity checks and gateway health. If you can reason about self-running agents and production cloud systems, you can pick up the specifics on the job.

This is not a research-only or pure backend role. You'll configure agents, write the code and docs, make product calls, and own decisions from day one.

Contract terms

• Duration: 3 months initial, with option to extend to 6 months by mutual agreement
• Work style: Remote-first, with a sustainable rhythm that protects working hours, leaves room for mentorship, and makes wellness part of the way we work
• Fixed component: ₹50k/month
• Profit share:10% of Saras's gross profit, paid at the end of the contract term, with no ceiling
• Tooling: All AI tools and subscriptions, fully expensed
• What could come next: If things work out between us and the business keeps booming, the role could convert to a founding engineer role with meaningful equity

After the 3-month period, we're open to renegotiate the terms of the contract subject to product trajectory and work performance.

What we're looking for

Expecting the following at a minimum, with some exceptions for stellar profiles:

• 2+ years of experience working on production-grade systems
• 1+ years of experience building/contributing to AI projects; exposure to OpenClaw/Hermes frameworks is a strong plus
• Fluency with concepts of agent memory, self-continuing loops and agent skills
• Above all, apply only if this resonates with you strongly: our manifesto — be prepared to ask and answer questions!

Applicants with compelling GitHub profiles or portfolios will be preferred over swanky resumes.