Senior full-stack and applied AI engineer

I build AI product systems that hold up outside the demo.

I have 10 years building product systems across web apps, ERPs, video-streaming platforms, mobile apps, backend services, analytics, user feedback loops, and growth campaigns. At OTTPlay, I led mobile work at 10M+ download scale; my current work adds applied AI on top of that base: WittyKeys, AI-SDR, RAG, and SFOS, the operating system I use for AI-orchestrated development, automated QA, evals, logs, and release confidence.

Career arc

Not a fresh AI pivot. A senior engineering base pointed at AI products.

2015-2017

Native Android foundation

Java/Kotlin, Play Store release cycles, mobile reliability, and platform constraints.

2017-2024

Full-stack product delivery

React/Next.js web apps, React Native and Android apps, Node/NestJS backends, ERPs, OTT/video streaming, logistics, on-demand platforms, MongoDB/MySQL, AWS, and direct founder/product collaboration.

2024-2026

OTTPlay leadership

Led a 5-engineer mobile team at 10M+ download scale via Wrocus / HT Media. Improved playback reliability, native player performance, React Native rendering, release quality, and product iteration from real usage signals.

2024-now

Applied AI + SFOS

WittyKeys, AI-SDR, Support Brain, Campaign Brain, automated QA, evals, and the solo-founder operating system around them.

Live Android product Claude + Firebase Billing + quota

WittyKeys is where the AI work became product work.

An Android AI keyboard and writing assistant for bilingual/Hinglish communication. The real work was not only prompt calls. It was onboarding, permissions, keyboard trust, overlay behavior, entitlement states, quota handling, release checks, logs, and Play Store polish.

Selected systems

Different projects, one through-line: useful AI needs a workflow around it.

OTTPlay
10M+downloads
11% -> 1.5%playback failure
Production leadership

OTTPlay mobile

Led cross-platform mobile delivery, native player work, streaming reliability, React Native performance, observability, sprint planning, code review, and stakeholder communication.

Read case note
AI-SDR campaign dashboard
Deployed / human-reviewed

AI-SDR

Campaign setup, prospect research, Claude-assisted drafting, regeneration, approval queues, follow-up, analytics, and browser E2E checks. Final sends stay human-approved.

Read case note
Support Brain knowledge gap dashboard
Development-stage

Support Brain

A RAG support assistant pattern with ingestion, retrieval, citations, confidence scoring, escalation, gap analytics, and support operations dashboards.

View links
Campaign Brain A/B testing dashboard
Demo-seeded

Campaign Brain

Analytics-to-action workflows for campaign calendars, recommendations, live performance, A/B testing, and optimization reports.

View links
SFOS

SFOS is the operating system I keep evolving around my work.

AI-orchestrated devCowork plans, Claude Code implements, parallel worktrees isolate work, and I approve the gates.
Automated UI workHTML references, implementation instructions, screenshot capture, visual comparison, and fix loops.
QA automationGolden screenshots, PixelDiffComparator, Espresso/UI Automator, real-device regression, and release checks.
AI quality loopsLLM-as-judge review, reply-quality scoring, trace design, and regression habits for outputs that drift quietly.
Operational memoryJourneyTracer, sanitized traces, logs, dashboards, docs, and post-release checks keep future debugging grounded.
Next role

Looking for senior applied AI, AI product, full-stack product, founding engineer, or forward-deployed roles.

I am strongest where product taste, full-stack execution, web/mobile systems, analytics, growth loops, and LLM workflow judgment all matter at the same time.