cdwip
Codedocket
A persistent developer context and decision system written in Go: typed knowledge entries, atomic writes, Git-backed history and provenance tracking, designed for coding agents with MCP integration in mind.
GoCLIGit
$ whoami
Hi, I'm Amaan.
I build systems that stay up.
I ship backend systems and cloud infrastructure that run live in production, co-author AI research, and win the occasional national hackathon. The parts users never see? That's where I live.
█
KHANTRONpower
go✦python✦typescript✦aws✦azure✦terraform✦docker✦postgresql✦dynamodb✦redis✦pytorch✦react✦next.js✦node.js✦fastapi✦linux✦git✦go✦python✦typescript✦aws✦azure✦terraform✦docker✦postgresql✦dynamodb✦redis✦pytorch✦react✦next.js✦node.js✦fastapi✦linux✦git✦
Experience
- Designed backend services for a domain-oriented microservices platform with independently deployable services.
- Built REST APIs and data integrations with Node.js/TypeScript, PostgreSQL and DynamoDB; auth via Amazon Cognito.
- Managed AWS infrastructure with Terraform and Serverless Framework: Lambda, API Gateway, IAM, SSM, Docker.
Node.jsTypeScriptPostgreSQLDynamoDBTerraformAWS
Proof of work
✦live
Donna
An AI assistant deployed for Riverbridge. Connectors ingest activity from Slack, Jira and GitHub into an embedding pipeline with semantic retrieval, running live in production.
AIRetrievalVector DB
oawip
OpenAgentRuntime
A lightweight runtime that lets AI agents pause, release their compute, and resume later without losing state, plus dynamic placement across ephemeral and persistent workers.
KubernetesRuntimeAI Infrastructure
Publications
Published at ICCTWC 2026 · DOI
Toolbox
Go
Python
TypeScript
JavaSQL
Node.js
FastAPIgRPC
Docker
Linux
AWS
Azure
Terraform
Serverless
Nginx
PostgreSQLDynamoDB
Redis
PyTorch
scikit-learnXGBoost
React
Next.js
Vite
Tailwind CSS
GitAlong the way
- ★1st Place · Hack This Fall, National Level (Dec 2025)
- ★1st Place · ISTD Hackathon (Jan 2025)
About
Hey, I'm Amaan, a full-stack engineer from Mumbai working across backend systems and cloud infrastructure. I care about the parts users never see: the deployment pipeline, the auth flow, the data layer that has to keep up for years.
Where I'm headed: AI infrastructure. I've co-authored research on optimal transport reward shaping and conformal retrieval, and I build the ingestion, retrieval and deployment plumbing AI applications actually run on; the goal being to work as an infrastructure AI engineer, one system at a time.
$ status
- location · Mumbai, India
- role · Full-Stack Engineer
- focus · backend, cloud, AI infrastructure
- trajectory · infra × AI
- availability · open to work