Backend engineer building distributed systems and ML-powered backends.

I design distributed systems, scalable microservices and ML-powered backend applications, and I build them end to end: the API, the database, the queues, the deployment and the observability. Rust is my main language, Python and Django sit alongside it, and AI coding agents are part of my daily workflow. I measure before I optimise.

Open to backend / SDE roles - remote or relocating.Lucknow, India

Work

Dec 2025 - Present

Backend Engineer (Contract)

Stealth startup, US (remote)

Backend and infrastructure for a pre-formation US startup, owned end to end. The product is under NDA, so this stays at the stack and work level.

Architected and deployed the core backend infrastructure from scratch: Rust (Actix-web), Railway, PostgreSQL and Cloudflare R2.
Designed the Tauri desktop app in Rust and TypeScript to render complex 3D scenes and 2D images, targeting 60 FPS.
Evaluated and integrated ML models in Python for complex sub-problems, optimizing accuracy with agentic AI (Claude, OpenCode).
Redis queues driving an event-driven architecture.
Designed an ML benchmark and accuracy-measurement tool in 4 days to compare pipeline speed and accuracy across versions, driving a throughput improvement.
Hosted the end-to-end stack on local dev environments with Bash, Docker and Compose across Linux and macOS, cutting dev-hosting costs.

Nov 2024 - Nov 2025

Backend Engineer

Amalaan India - BugFlows

BugFlows is an AI enterprise bug-triage platform that predicts defect assignment, similarity, ownership and resolution time. I designed and ran the backend and infrastructure.

Built the core Rust (Axum) REST API server from scratch in about 2 months: authentication, APIs, PostgreSQL, job orchestration, streaming uploads and ML service integration.
Held the service to sub-50 MB of memory even while handling 1 GB+ multipart CSV uploads, by streaming to disk directly.
Designed a Rust task scheduler to start, stop and auto-scale expensive ML VMs across GCP and AWS - 70% cost savings on VM bills.
Built the REST APIs in Python Django and the integrations with GCP Buckets and AWS S3 that store and upload defect data; scaled the architecture to process 100K+ defects from GitHub, YouTrack and Bugzilla.
Dockerized the stack: Rust backend, Postgres, Django ML service, Caddy and the Grafana/OpenTelemetry stack (Alloy receiver with Loki, Prometheus and Tempo sinks), with image builds on a dedicated GCP build VM.

Feb 2021 - Aug 2023

Programmer Analyst

Cognizant - client: Swiss Re

Enterprise applications for Swiss Re, the world's second-largest reinsurer, in C#/.NET, SQL Server and MS Access.

Owned the Excel integration of a C#/.NET WPF app used org-wide at Swiss Re, reaching 92-95% feature parity; also worked on the app's Word and PowerPoint integrations.
Optimized SQL queries and cut the execution time of a complex query to under 10 seconds from over a minute.
Designed an internal tool that compares two MS Access databases and generates a single color-coded report, replacing manual review of 10+ tables.
Joined through Cognizant's graduate training programme (Programmer Analyst Trainee): C#, .NET, SQL Server, AngularJS, Docker, Linux and Kubernetes.

Projects

GTF Field-Survey App

Flutter · Firebase · Google Maps

A field-survey app built during a Ministry of Environment, Forest and Climate Change internship.

500+ surveyors recorded flora, fauna and human-activity sightings along the Ganga, with custom sighting types on a map.

Fully serverless: the app talked straight to Firebase for auth, database and storage. Three versions shipped; over a year in field use.

Google Foobar (2020): finished all five problems in the invitation-only Google programming challenge.

About

I have spent about four years on backend work: first on enterprise applications for Swiss Re at Cognizant, then building the BugFlows product backend from nothing, and now the backend for a stealth startup in the US.

The through-line is that I like knowing how things actually work, and that measuring beats guessing: the benchmark tool came before the throughput improvements, and the cost numbers came from the bills.

Right now I am building the backend for a stealth US startup and looking for the next backend or SDE role: remote, or relocating.

Languages: Rust, TypeScript, Python, C#, SQL
Backend: Tokio, Tauri, Axum, Actix-web, Django, REST APIs, Redis
Cloud & Data: AWS, GCP, PostgreSQL, Cloudflare R2
Infrastructure: Docker, Linux, Bash, Git, CI/CD, GitHub Actions, Caddy, Railway, RunPod
Observability: OpenTelemetry, Grafana, Prometheus, Loki, Tempo, Grafana Alloy
AI & LLM: Claude, OpenCode, DeepSeek, ChatGPT

Contact

If you are hiring, or just want to talk shop, I would like to hear from you.