I build systems that survive contact with real users.
I'm Lalit — an engineer working at the intersection of AI, full-stack development, and marketing automation. I prefer shipping over speculating, and I treat every project as something a future engineer will inherit.
AI Engineer · SFMC Developer · Full Stack Builder
India · Working remotely worldwide
Full-time roles · Select freelance engagements
How I got here.
I started in a fairly traditional path — an MCA, the usual coursework, the usual expectation that "developer" meant "writes Java for someone else's enterprise app." That stopped fitting almost immediately.
The pivot happened in two places at once: building real products for myself (BrainBench, ThreadTalk, ProjTrack Desk) and getting hands-on with AI infrastructure at CloudCove.ai. The combination was the lesson — products teach you what users actually need; industry teaches you what production actually demands.
Today I split my time between AI engineering work — RAG, LLM orchestration, agent systems — and the full-stack product work that surrounds it. I take fewer projects than I could, on purpose, because I want every one to ship something I'd be willing to put my name on.
The path, in four chapters.
- 2022
Started MCA
Began a Master's in Computer Applications with a clear bias toward systems thinking — databases, distributed work, and the engineering side of CS rather than the theory.
- 2023
First serious products
Built BrainBench and ThreadTalk — the first projects that taught me what 'shipping' actually means: edge cases, deployment, real users, real bugs.
- 2024
Industry exposure — CloudCove.ai
Worked alongside engineers shipping production AI infrastructure. Learned what 'good' looks like at scale: evals, observability, and the discipline of saying no to scope.
- 2025
AI engineering, full-time focus
RAG systems for academic institutions, ProjTrack Desk in production, and a growing freelance practice for teams who need an engineer who can hold the full picture.
The tools I reach for.
Not exhaustive. These are the tools I've shipped production work with and would use again tomorrow.
"AI should disappear into the product. The user should just feel it works."
I'm betting the next decade of great software won't look like chatbots bolted onto dashboards. It'll look like products that quietly understand context, make fewer demands of their users, and earn trust through reliability — not novelty.
Let's build something