Sarang Pramode
Machine learning and backend engineer working on LLM-powered systems that need to behave reliably in production.
Where I focus
I'm drawn to systems that have to hold up in the real world, especially when reliability, observability, and scale actually matter. A lot of my work lives in the gap between a promising model and a production system people can depend on every day.
I care a lot about the parts of systems that usually only get noticed when they break: responsiveness, safety and compliance, clear failure surfacing and traceability, performance tuning, retrieval optimization, and robust tool integration. I've spent a lot of time in that messy layer between models and production, and I like turning it into systems that feel calm, dependable, and easy to trust as they evolve.
I'm especially interested in augmenting the real world with intelligence — using technology to enable people to thrive. I'm actively exploring how to build environment-aware, intelligent systems that are practical, reliable, and truly impactful.
Stack & Skills
- Machine learning & LLMs
- PyTorch, TensorFlow, Scikit-learn, Hugging Face, LangGraph, ADK, RAG, evaluation pipelines, guardrails, red teaming
- Backend & data systems
- Python, SQL, FastAPI, APIs, ETL pipelines, Pandas, product analytics, A/B testing
- Infra & platform
- AWS, GCP, Docker, Kubernetes, Helm, Redis, Linux, Git
- Observability & MLOps
- MLflow, Langfuse, OpenTelemetry, Grafana, Splunk, model monitoring, experiment tracking, CI/CD
Now
- Thinking about and what our interfaces with technology could look like
- Writing about Tech, Systems, and cool things I find on the internet