How I built an engineering team, three AI products, and a Swiss enterprise client base — from the first line of code
Ethereal Covenant — Bengaluru, India I was the first hire. There was no team, no codebase, no product. Over two and a half years I built the engineering foundation, shipped products for international clients, and grew the team from zero to ten.
Client:
Halo
Duration:
2 Years 8 Months
Location:
Bangluru, India
Role:
Lead Product Engineer → Engineering Manager
The Problem
Ethereal Covenant had a vision — build AI-powered products for real business problems. What they didn't have was an engineering team, an architecture, a workflow, or a codebase. I joined as the first engineer with a mandate to build everything from the ground up while simultaneously delivering for paying clients.
What I did
For Nxgen Technologies AG (Switzerland)
Nxgen builds NXG GENESIS — a cloud platform connecting security, fire prevention, IoT and AI for enterprise customers across Germany, Switzerland, Poland and Benelux. They needed a white-label multi-tenant mobile platform that their enterprise clients could deploy under their own brand.
I built it. The result: Nxgen grew from 1 enterprise client to 7. A Swiss security company with serious international customers — Hanwha Techwin, Securiton, Securitas — now had a mobile platform that could scale across their entire client base.
For SpeakUp Now (USA)
SpeakUp Now is an AI-enabled SaaS platform for workplace culture — built around encrypted, confidential communication between employees and organisations. Their platform had grown organically and the codebase was becoming a bottleneck — adding features was slow, deployments were risky, and teams were stepping on each other.
I integrated 14+ micro-frontends across their platform — modularising the architecture so that features could be developed, tested, and deployed independently. The result was faster delivery, cleaner code, and a platform that could scale without grinding to a halt.
Lytespeed.ai
Built the company's own flagship AI product — a no-code workflow builder for deploying AI voice agents across 22 Indic languages. Zero to live SaaS in 4 months. Full story in the Lytespeed case study.
The engineering team
Scaled from 0 to 10 engineers — hired, onboarded, established coding standards, set up agile workflows, and transitioned into the Engineering Manager role as the team reached a size that needed dedicated leadership.
Result
Three products shipped for international clients. One internal product launched. One engineering team built. Nxgen went from 1 to 7 enterprise clients on the back of the mobile platform. SpeakUp Now got a modular architecture that could actually scale. Lytespeed launched as a live SaaS product in six months.
Stack & Tools
