Backend engineer in Bengaluru, India. I write Go and Python, keep state in PostgreSQL, and run things on AWS. Most of my work is the unglamorous kind — job queues that retry properly, APIs that don't leak internal IDs, pipelines that survive a bad input at 3am.
At Indee I work on the video system: transcoding pipelines on AWS Elemental and Zencoder, delivery and playback, retention jobs across 4.5M+ assets. I own a core platform module end to end — APIs, features, CMS integration — plus the subscription billing on top of it.
Outside work I'm going deeper on Go (runtime, concurrency, tooling) and on AI agents — RAG pipelines, multi-agent orchestration, and how you actually evaluate any of it.
| Languages | Go · Python · JavaScript |
| Backend | Django · FastAPI · Flask · Gin · NodeJS |
| Databases | PostgreSQL · MongoDB · Redis |
| Cloud | AWS (EC2, Lambda, S3, Elemental, DynamoDB) · Docker |
| DevOps | CI/CD · OpenTelemetry · Prometheus · Grafana |
| AI/ML | LangChain · HuggingFace · TensorFlow · RAG |
| Systems | REST · gRPC · WebSockets · SSE · Microservices · Celery · RabbitMQ · FFmpeg |
prrev — self-hosted multi-agent PR reviewer. RAG over your
codebase with pgvector, River for job orchestration, OTel spans so you can see why it said something
dumb. Slack bot and VS Code extension. go postgres pgvector river
videovault — file encryption that hides data inside
playable MP4s. AES-GCM with Argon2id, payload written into frames via LSB steganography. The output
is a real video. It plays. go cryptography
Pratyaksh — real-time infrastructure progress tracking using GIS and image processing. Won Smart India Hackathon 2023. demo
LearnZ — e-learning platform with an NLP recommendation
engine that maps an uploaded syllabus to relevant lectures. python django docker
inpycon2025 — contributed to the PyCon India 2025 website, an open-source conference with 1500+ attendees.
sgagan.dev · resume · blog · @Astrax_04 · linkedin · sgagan2542@gmail.com