Home About Prashant

Prashant, founder of The Code Smith

Prashant.
Founder & builder.

The person who reads your brief is the person who writes the code. No account managers, no hand-offs — fifteen years of building and running SaaS, applied directly to your project.

01

Fifteen years of shipping

I've spent 15+ years building SaaS products — the front-ends people click, the APIs underneath, the databases, the queues, the billing and auth nobody sees until it breaks. The Code Smith is my studio, and it's deliberately small: one accountable builder rather than a pyramid of hand-offs. When you brief a project here, you're not being routed to a junior team learning on your invoice.

Experience — 15+ years, SaaS
You brief — the person who builds
02

I run my own products

I don't just build software for clients — I build and operate my own, on exactly the stack I'd put you on. LeadProwl, a lead-data platform. NoMailBounce, an email-verification service. Unbuilt Lab, a startup-idea research app. Plus a library of nineteen live website templates and a self-hosted RAG chatbot. These aren't demos — they're production systems I keep running, which means the advice you get comes from operating scars, not conference slides.

Products I run — LeadProwl · NoMailBounce · Unbuilt Lab
Also — 19 live templates, a self-hosted chatbot
03

How I build

Hand-written code — no page builders, no bloated themes, no black boxes you can't inspect. I work in Python, FastAPI and Postgres on the backend; React and React Native / Expo on the front; queues and schedulers for the work that runs around the clock; and self-hosted LLM / RAG where AI genuinely earns its keep. The through-line is ownership: I architect things so you end up owning your stack outright — code, docs and credentials — with zero lock-in.

Stack — Python/FastAPI/Postgres · React · React Native/Expo
Also — queues, schedulers, self-hosted LLM/RAG

What I write about

Lessons from running the thing.

Everything I publish comes from work that's actually in production — no theory I haven't lived with.

AI in production

Where LLMs and RAG genuinely pay off, where they quietly don't, and how to keep them factual and grounded once real users arrive.

Automation & SaaS

Turning repetitive work into pipelines that run themselves, and shipping honest MVPs before over-building.

Ownership & craft

Hand-written code, own-your-stack architecture, honest deliverability, and why lock-in is a choice you don't have to make.

Have a bottleneck?
Brief me directly.