I build software with a focus on backend systems, APIs, integrations, performance, and practical architecture.

I'm currently exploring AI engineering through hands-on projects — from building RAG applications to understanding how LLMs work from the ground up.

• Backend Development
• Software Architecture & Design
• API & System Integration
• Performance Optimization
• Healthcare Software
• AI & RAG Applications
• Cloud & Distributed Systems

Backend

• .NET / ASP.NET Core
• C#
• Python
• FastAPI
• Node.js

Frontend

• React
• TypeScript
• JavaScript

AI / LLM

• RAG
• Embeddings
• Vector Search
• Ollama
• LLM Architecture
• Transformers

Cloud & DevOps

• AWS
• Azure
• Docker
• Git

Database

• SQL Server
• PostgreSQL
• MySQL
• Redis

A provider-agnostic Knowledge Assistant built with FastAPI, React, PostgreSQL and RAG.

The project covers:

• Document processing
• Document chunking
• Embeddings
• Vector search
• Semantic retrieval
• Grounded AI responses
• LLM provider architecture
• Background processing
• Full-stack integration

View Project →

A hands-on project to understand how modern language models work by building the fundamental pieces from scratch.

Learning path:

Neural Networks → Backpropagation → Tokenization → Embeddings → Attention → Transformers → Language Model

View Project →

I document practical engineering lessons, decisions, problems, and things I've learned while building software.

Topics include:

• Architecture
• Performance
• Databases
• Entity Framework
• Raw SQL
• Caching
• Authorization
• Production troubleshooting

• When Should You Use Redis Cache?
• Why Do Entity Framework Queries Become Slow in Complex Reporting Systems?
• When Should You Use Raw SQL Alongside Entity Framework?
• How Do You Troubleshoot High Database CPU Usage in Production?
• How Do You Implement Data-Level Authorization?

Read the Engineering Journal →

• LLM fundamentals
• AI application architecture
• RAG systems
• Local LLMs
• AI coding agents
• Modern software architecture

• Understand the problem before choosing the technology
• Keep solutions simple
• Design before coding
• Measure before optimizing
• Build things that are easy to understand and maintain
• Never stop learning

• LinkedIn: https://www.linkedin.com/in/anparmar/
• GitHub: https://github.com/parmaramit1111