I build software that respects the people who use it.

R&D Software Engineer with 10+ years of experience designing scalable, resilient systems on cloud-native platforms. Most of my work orbits a few core beliefs: your data should stay yours, the best features don't need a server, and complexity should live in the architecture, not in the user's face.

TL;DR — Java/Spring Boot engineer, 10+ years, cloud-native & Kubernetes. Build privacy-first apps and dev tools on the side, shipping them under arcanelabsio. Tinker with local LLMs and on-device AI. ISB Emerging Leaders alum. LinkedIn.

• Platform engineering — build and operate Kubernetes-based platforms that integrate with resource providers, support advanced release strategies (red-black, canary, rolling), and right-size resource requests based on actual consumption to eliminate over-provisioned, under-utilized infrastructure
• Control plane engineering — runtime-tunable observability, CLI tooling for platform APIs, health monitoring across control and data planes for zero-downtime deployments
• API strategy — designed API hub patterns (discover, develop, distribute) for microservices ecosystems
• Mentorship — guiding cross-functional teams, mentoring engineers, and building a culture of ownership and continuous learning

All shipping under arcanelabsio — the umbrella for privacy-first, local-first software I build in the evenings.

Longeviti — A bring-your-own-AI longevity tracking app. Your blood work and daily logs drive adaptive meal and supplement plans — but the app itself does zero AI compute. Your own Claude or Codex session reads your reports, writes plans that conform to public JSON schemas, and lands them on your own Google Drive. No servers, no inference fees to us, no data we can see. Docs · Framework.

Vael — A privacy-first family finance app for Indian households. Local-first, encrypted, zero-cost. Families get a shared financial picture — net worth, budgets, goals, projections — without surrendering data to anyone. Built with Flutter, Drift ORM, AES-GCM encryption, and Google Drive as a dumb encrypted pipe. 400+ tests, TDD from day one.

Forge — A CLI that gives AI coding assistants superpowers over your GitHub workflow. One install plugs into Copilot, Claude Code, Codex, and Gemini CLI — PR reviews, discussion analysis, commit coaching. Zero lock-in.

Shellcraft — A macOS dev machine bootstrapper with profile-based package selection and non-destructive dotfile management.

• Privacy as architecture — not a toggle, not a policy. Encryption boundaries, client-side-only processing, cloud providers that see only noise.
• Local-first — apps that work on an airplane, cost nothing to run, and don't depend on someone else's uptime.
• Deterministic over intelligent — clear math and rules over black-box predictions. Users make judgments; tools give them accurate numbers.
• Distributed systems — love the hard problems: orchestration, consistency, failure boundaries. Comfortable with tools like Temporal for durable workflows.
• Developer tooling — making the feedback loop tighter, the workflow smoother, the assistant smarter.

Core (write it cold): Java · Spring Boot · Golang · Python · Kubernetes · Docker · Microservices · Kubebuilder · Temporal · MySQL · PostgreSQL · MongoDB · Redis · S3 · GitHub Actions · Jenkins · CI/CD

AI-assisted (strong process knowledge, build with AI pairing): Flutter · Dart · TypeScript · Node.js · SwiftUI I understand app development end-to-end — TDD, E2E testing, simulator workflows, app distribution, clean architecture — and use AI as a force multiplier to ship production-quality code in these stacks.

Certifications: Pivotal Certified Spring Professional · Oracle Certified Java Associate

Education: ISB Emerging Leaders Program · B.E. (IT), Sastra University

Core Java developer by trade — equally comfortable rolling up sleeves with UI teams to ship end-to-end.

I'm drawn to solving everyday friction with software — the kind of problems that aren't glamorous but save real time:

• File deduplication — taming years of scattered backups and duplicates across drives
• Speech diarization — on-device transcription that knows who said what, no cloud required
• Local LLMs — exploring what runs well on Apple Silicon and how to wire it into daily workflows
• AI plugins for life and work — nothing invasive, just quiet automation that compounds into 100x productivity

The thread across all of it: tools that run locally, respect your privacy, and remove friction you didn't know you had.

_{I'm interested in the intersection of personal software, financial sovereignty, health data ownership, and on-device AI. If any of that resonates — let's connect.}