Going solo in 2025

There’s a particular kind of clarity that comes after you’ve been part of building something from nothing. You understand how early teams work, what founders actually need, and, more importantly, what you’re capable of when the stakes are real.

That’s where I was at the start of 2025. I had spent the better part of the previous year as one of the earliest engineers at a software startup: internal systems, SaaS products, mentoring teammates, and even working embedded within a partner company. By the time that chapter closed, I had enough signal about how I work best: close to the problem, close to the founders, and with genuine ownership over what I build.

So in February 2025, I made a deliberate choice. No job boards. No mass applications. I went deep into the startup ecosystem, and built something of my own while doing it.

That decision led me to formally register Sravan Tech Labs, my product-based software engineering practice. What started as independent collaborations became a proper business: B2B arrangements with startups as a business owner, not just a hired hand. Every engagement has been company to company: more accountability, more ownership, and a different kind of trust on both sides.

Why now

Early-stage startup work teaches you something hard to articulate until you’ve lived it: the best work happens when the person building actually cares about the outcome, not just the ticket or the sprint, but the outcome.

Starting Sravan Tech Labs wasn’t about leaving structure behind. It was about building the right structure, one where I could move fast, make real technical decisions, and work directly with founders solving problems worth solving. Formalising it meant startups weren’t just hiring a developer; they were partnering with an engineering practice that had skin in the game.

I wasn’t looking for clients. I was looking for collaborators.

Koramangala

If you’re in Bangalore and anywhere near startups, you know what Koramangala means: density of founders, builders, and early-stage companies. I started spending time there: meetups, coffee, listening more than pitching. Casual conversations turned serious. Founders shared real problems; I shared how I’d approach them. Those conversations became partnerships.

Most of my 2025 work began not through a proposal or a platform, but through a conversation that went somewhere unexpected.

Illumora

Illumora is an early-stage startup in mental wellness, helping people get better, not just feel better. They had a clear vision and a real product, but their web presence wasn’t doing justice to either.

I built a multi-domain web presence covering the core brand and sub-brand Boombooth Studio, personal branding and podcasting for individuals building voice and presence. Two audiences, two tones, one coherent technical foundation.

Wellness and personal branding are deeply personal categories. Every layout and interaction carries more weight than on a generic SaaS product. That work sharpened instincts for design and communication beyond pure engineering.

Communn

Communn builds digital solutions for early-stage businesses that are community-first: local owners, emerging entrepreneurs, and small teams who need POS, web presence, and operational tooling without enterprise budgets.

I helped translate their vision of accessible digital infrastructure into working product, community tooling, web presence, and the stack that supports clients day to day. Users aren’t developers; they care whether it works when they need it. That keeps product decisions honest.

Chord AI

This was product work in the truest sense: not a single feature, but full-stack responsibility across the lifecycle. I audited design vs usability, then built frontend, backend, and deployment end to end.

The hardest piece was multi-user flow at scale: roles, permissions, and expectations of the interface. Making it work reliably meant rethinking architectural assumptions. There’s “full-stack” that touches both layers, and “full-stack” where you own design decisions, deployment, and failure modes. Chord AI pushed me firmly into the second.

Agrigrader

I met the founder the same way I’ve met my best collaborators this year: no agenda, just a conversation. Agrigrader is substantive agro-tech backed by nearly a decade of domain experience: a multi-vendor, multi-user marketplace on web already at scale.

I’m contributing to their mobile app: a multi-vendor experience that extends the web platform. The complexity is respecting and surfacing a sophisticated web-first backend while delivering something native on mobile. Extending an existing product demands deep understanding of what you can innovate vs what you must keep faithful to.

The practice

Formalising the work was the natural conclusion of 2025. It’s a registered software engineering practice: product-focused, startup-native, structured as B2B. Every startup this year engaged as a business partner, not as an individual contractor.

When two businesses collaborate, the conversation shifts from “here’s a task” to “here’s a problem: how do we solve it together?” That’s the kind of work I’m built for. Every founder came from a conversation, not a contract first. Every product I’ve touched is live, used by real people.

Sravan Tech Labs is open to select collaborations with early-stage startups. If you’re building something interesting, let’s talk.