About
OnFinance AI · previously Swiggy and Karta · IIT Bombay, B.Tech 2026
I'm a forward deployed engineer at OnFinance AI Peak XV Surge · $4.2M in Mumbai. The job sits between a product team and a client's engineering, security, and risk teams: I scope what an institution actually needs, build it on top of our agent harness, deploy it inside their environment, and stay responsible for it after launch.
What "forward deployed" means to me
Banks don't run your software in your cloud. They run it on their hardware, behind their network, next to core banking systems that predate most of the people maintaining them. A forward deployed engineer is the person who makes an AI system work under those constraints and can explain, to a risk committee, why its outputs can be trusted.
In practice that means three kinds of work. Building the harness: orchestration loops, tool interfaces, context assembly, and deterministic fallbacks so that every output is traceable. Deploying it: on-prem and hybrid rollouts, legacy integration, holding latency and reliability budgets. And keeping it honest: continuous evaluation and tracing against live traffic, with regression checks that gate every release.
I wrote a longer version of this, including who the role doesn't suit, in Forward Deployed Engineer: What the Title Doesn't Tell You.
Background
I studied Metallurgical Engineering and Materials Science at IIT Bombay (2022 to 2026). The route into AI came through research: my thesis fine-tuned a protein language model to design antimicrobial peptides, which taught me more about data curation and evaluation than any course did. Along the way I interned at Swiggy, building an enterprise RAG system over internal knowledge bases, and at Karta $3M seed · Oct 2026, building an observability platform on LangSmith traces for agent workflows.
Outside work
- Convener of SilverScreen, IIT Bombay's film and media club. Two films from Inter IIT Cultural Meet 6.0: Talaab (gold, 51-hour filmmaking competition) and Paithani (bronze, online filmmaking competition).
- Team lead at ShARE IITB.
- Granted L2 funding by IDEAS, IIT Bombay's innovation programme, for a student project.
Contact
The fastest way to reach me is ajinkyamawal.iitb@gmail.com. I'm also on LinkedIn and GitHub, and there's a one-page resume.