A practice built on thinking before building.

I’ve spent the last seventeen years building production SaaS — most of that as the technical lead, latterly as the founder and CTO of Bhavi Technologies, a small engineering studio based in Vadodara, India.

I started Bhavi in 2012 because I wanted to build a service company that helps global clients ship quality products and stay supported for the long term — not a body shop, not a delivery factory. The studio worked with European clients for over a decade. Today most of my time goes into a long-running European engagement where I lead architecture for their core platform, and into building three of my own products: Cloudbrief and Paperbrief (the two AI products under Kashi Labs), and Employrax.

The work I’m proudest of isn’t the code. It’s the architectural calls that hold up three years later — the multi-tenant boundary that prevents the data leak nobody is testing for, the modular monolith that lets a small team ship without microservice overhead, the compensation snapshotting that means historical payroll never silently changes. Most software fails on decisions made in the first month, not the last. My job is to make those decisions count.

The right time to think about scale is during the data model design — not after the third outage.

Multi-tenant data isolation is structural, not careful. If your safety depends on a developer remembering to add WHERE tenant_id = ? to every query, you’ve already lost.

Microservices solve organizational problems, not technical ones. Most early-stage products that adopt them regret it within two years.

AI-assisted coding hasn’t made architects obsolete. It’s made architects more valuable, because the cost of writing bad code dropped to zero while the cost of designing the wrong system stayed exactly the same.

Boring stacks win. PostgreSQL, Laravel, FastAPI, React, AWS — used with discipline — outlast every framework-of-the-month.

I design first, then build. Specifications, architecture diagrams, and rollout sequencing come before any line of production code. This is the part that distinguishes a senior engagement from a freelance one: clients pay for the decisions they don’t have to make.

For implementation, I work the way most senior engineers do in 2026 — directing modern AI coding tools (Cursor, Claude) under tight supervision rather than typing every character by hand. The architecture is mine; the velocity comes from the tools. Every line that ships, I review and own. If something breaks at 2am, I’m the one who fixes it.

I work with a small number of clients at a time — usually one or two active engagements — so each gets real attention.

I don’t interview, take coding tests, or “prove myself” through trial tasks. After seventeen years, the proof is the work — the products I’ve built, the architecture I can walk you through, the live systems you can look at. If you want to evaluate me, the right way is a conversation about your problem, not a take-home exercise.

What I offer instead is straightforward: a call where we discuss what you’re trying to build, I ask the questions a senior architect would ask, and we both decide honestly whether I’m the right person to help. If I’m not, I’ll tell you and often I can point you to someone better. If I am, we agree on scope and start.

I take engagements that are well-defined enough to do good work in. I turn down projects where I’d be a commodity coder, where the spec is “just build what we tell you,” or where the buyer wants seniority but isn’t willing to pay for the thinking that produces it.

The work I do happens through three brands I run today: Kashi Technology Pvt. Ltd., the registered parent entity in Vadodara; Kashi Labs, the AI-product brand under which Cloudbrief and Paperbrief ship; and Krayons Technology, my consulting practice. Together they’re the current chapter of a longer journey.

I’ve been building software for clients under one banner or another since 2008, and running my own companies since 2012. Each transition has been about getting the structure right for what I wanted to build next.

— 2012 — Founded Bhavi Technologies India Pvt. Ltd. as a service company for global clients. A decade of long-running European engagements, mostly architecture and delivery for SaaS platforms. Dissolved in 2023.

— 2023 — Started Krayons Technology as an independent consulting practice, refocused around a smaller portfolio of senior architecture and AI-integration engagements.

— 2026 — Founded Kashi Technology Pvt. Ltd. as the registered parent. Stood up Kashi Labs as the AI-product brand and shipped two products under it — Cloudbrief and Paperbrief. Krayons operates under Kashi’s umbrella; Employrax sits directly under Kashi Technology.

Cloudbrief — a daily AI brief on your AWS account. Read-only IAM onboarding, deterministic detectors, Claude synthesis only when signals corroborate. Live at cloudbrief.app under Kashi Labs.

Paperbrief — a multi-tenant RAG document Q&A platform with per-tenant isolation enforced in PostgreSQL. Originally shipped as DocSense; now live at paperbrief.app under Kashi Labs alongside Cloudbrief.

Employrax — an India-first HR and payroll SaaS designed for European SMBs. Multi-tenant from day one, with statutory compliance (TDS, PF, ESI, Form 16) as a first-class capability. Sits directly under Kashi Technology, separate from the AI-product brand.

I take on one or two new engagements at a time. Most work falls into one of four shapes: an AI integration audit, a multi-tenant SaaS architecture review, a fractional architect retainer, or a defined build project. See the services page for specifics.

If you’ve read this far and something I’ve described matches what you need, I’d be glad to talk.