# Niranjan Sah - Full Stack Developer

> Full Stack Developer specializing in React.js, Next.js, FastAPI, and scalable web applications. Building modern software with modern technologies.

Niranjan Sah is a full-stack software engineer based in Mirchaiya, Nepal, working across frontend, backend, AI systems, and cloud infrastructure. Available for freelance and remote work, and the founder of Niranjan Works, an independent software studio.

## Key facts

- Name: Niranjan Sah
- Role: Full Stack Developer · Freelance Software Engineer · Founder of Niranjan Works
- Location: Mirchaiya, Nepal (open to remote and relocation)
- Studio: [Niranjan Works](https://niranjanworks.com) - full-stack development, AI & intelligent systems, product engineering, cloud & infrastructure
- Core stack: React, Next.js, TypeScript, FastAPI, NestJS, Node.js, PostgreSQL, Redis, Docker, AWS
- Primary site: https://www.niranjansah87.com.np

## Hire / work with me

Open to freelance projects, contract work, and full-time remote roles. Services available through Niranjan Works: web applications, SaaS platforms, APIs & integrations, admin dashboards, AI integrations, and cloud infrastructure.

- Studio: https://niranjanworks.com
- Start a project: https://niranjanworks.com/start-a-project

## Profile

- [About](https://www.niranjansah87.com.np/about.md): A quick intro to me, my tech stack, and how to connect.
- [Experience](https://www.niranjansah87.com.np/experience.md): Highlights from my career and key roles I've taken on.
- [Projects](https://www.niranjansah87.com.np/projects.md): Selected projects that show my skills and creativity.
- [Certifications](https://www.niranjansah87.com.np/certifications.md): Certifications and credentials I've earned.

## Blog

- [What Auditing a Production AI Platform Taught Me About Engineering](https://www.niranjansah87.com.np/doc.mdx/auditing-production-ai-platform-lessons/): I spent two weeks systematically evaluating an AI platform's production behavior. What I found contradicted most of what the documentation claimed - and changed how I think about building systems.
- [Why Observability Comes Before Optimization](https://www.niranjansah87.com.np/doc.mdx/observability-before-optimization/): Every optimization I've seen fail started the same way: an engineer who was confident about where the problem was, before they'd measured anything. Here's what changed how I think about building systems.
- [Architecture Decisions Compound Over Time](https://www.niranjansah87.com.np/doc.mdx/architecture-decisions-compound-over-time/): Every architectural decision is a bet on the future. Most bets are wrong - not because the decision was bad, but because the world changed. Here's how small choices accumulate and how to make decisions that age well.
- [Secure Environment Manager: Designing a Self-Hosted Secrets Platform](https://www.niranjansah87.com.np/doc.mdx/secure-environment-manager/): How I designed and built a production secrets management system - the threat model, the architecture decisions, and the trade-offs I didn't anticipate.
- [When Documentation Lies](https://www.niranjansah87.com.np/doc.mdx/when-documentation-lies/): The most dangerous gap in engineering isn't bad code - it's documentation that used to be true. Here's what I've learned about documentation drift and how to prevent it from becoming a production hazard.
- [What 283 Production Requests Taught Me About Assumptions](https://www.niranjansah87.com.np/doc.mdx/what-production-requests-taught-me/): I analyzed 283 production API requests to understand how users actually interact with an AI platform. What I found contradicted several assumptions our team had made about how the system would be used.
- [The Hidden Cost of LLM Context](https://www.niranjansah87.com.np/doc.mdx/llm-context-management/): Every RAG system I've worked with had the same silent failure mode: retrieval that looked correct but wasn't. Here's what I learned about managing context quality in production AI systems.