AI & Backend EngineerComputer Vision • LLM Systems • Real-time AI

Building production AI systems that solve real problems - from driver safety monitoring to intelligent document processing.

View ProjectsGet in Touch

About Me

A snapshot of who I am and what I do

I'm an AI & Backend Engineer who builds production-grade infrastructure that brings AI to life. My focus is on transforming cutting-edge models into reliable, high-performance applications that solve real-world problems. From real-time computer vision to privacy-first LLM integrations, I architect solutions that are both intelligent and scalable.

My expertise covers RESTful APIs, WebSocket servers for ML model serving, async processing pipelines with RabbitMQ, and inference optimization with TensorFlow Lite and TensorRT. I've architected systems processing video at 30 FPS for drowsiness detection, RAG pipelines handling 45+ pages/min with LangChain, and offline voice assistants achieving <5s response with local LLMs. Whether it's FastAPI, PostgreSQL/MongoDB, or Docker on AWS, I focus on performance and scalability.

I build offline-capable, privacy-respecting systems optimized for edge deployment. I don't just integrate models – I architect the infrastructure around them, ensuring reliability under real-world conditions. My work bridges AI research and production engineering, creating backend systems that are robust, efficient, and ready for scale.

Karan Parekh

Key Highlights

AI & Backend Engineer with a focus on scalable, production-grade systems
Built real-time computer vision pipelines and privacy-first LLM integrations
Expert in FastAPI, TensorFlow, OpenCV, LangChain, and edge deployment
Architected async processing pipelines, RAG systems, and WebSocket streaming
Designed RESTful APIs and WebSocket servers for ML model serving
Optimized inference with TensorFlow Lite and TensorRT for edge AI
Built offline voice assistants with <5s response using local LLMs
Deployed scalable backends with Docker, PostgreSQL/MongoDB, and AWS

Location

Ahmedabad, India

Experience

3+ Years

Focus

AI & Backend

Featured Projects

Deep technical case studies showcasing architecture, challenges, and solutions

View all projects

Skills & Expertise

Technical skills and tools I use to build AI and backend systems

Programming Languages(3)

Python

expert

Backend services, ML/vision prototypes, FastAPI, data processing pipelines

TypeScript

expert

Backend microservices, NestJS for API services

JavaScript

advanced

Node.js services, backend development

Backend & Frameworks(5)

FastAPI

expert

High-performance Python APIs with async support, automatic OpenAPI docs

Node.js

expert

JavaScript runtime for scalable backend services, event-driven architecture

NestJS

expert

Service-oriented APIs, dependency injection, scalable architecture

RabbitMQ

advanced

Messaging between services, event-driven patterns

Kafka

intermediate

Event streaming for data ingestion pipelines, real-time processing

Cloud & Infrastructure(4)

AWS

advanced

S3 for video storage, boto3 SDK, cloud infrastructure

Azure

advanced

Azure AI Document Intelligence, Cosmos DB, Azure OpenAI Services

Docker

advanced

Containerization, multi-stage builds, image management

Google Cloud

intermediate

Vision OCR, Gemini API integration for document processing

Databases & Storage(4)

MongoDB

expert

Document store for embeddings, application data, NoSQL queries

PostgreSQL

advanced

Relational storage, complex queries, schema introspection

Cosmos DB

intermediate

Vector search for RAG systems, globally distributed NoSQL

OpenSearch

intermediate

Full-text search, analytics, log aggregation

Machine Learning & Vision(10)

TensorFlow & TF Lite

expert

Custom model training, TF Lite optimization for edge deployment

OpenCV

expert

Real-time video processing, facial landmark detection, computer vision pipelines

MediaPipe

expert

Face Mesh, Blendshapes, facial landmark tracking for drowsiness detection

Vision Transformers

advanced

ViT-based models for unified detection and pose estimation

YOLO

advanced

Object detection, real-time inference optimization

OCR & Document AI

expert

Google Vision OCR, Azure Document Intelligence, text extraction

LLM Integration

expert

Ollama, LangChain, local and cloud LLMs, prompt engineering

RAG & Embeddings

advanced

Vector embeddings, semantic search, retrieval-augmented generation

Speech Recognition

advanced

Whisper, faster-whisper, CPU-optimized STT for offline processing

Text-to-Speech

advanced

Coqui TTS, pyttsx3, multi-engine TTS for voice assistants

Real-time & Streaming(3)

WebRTC

advanced

Camera streaming from Raspberry Pi, mobile access via hotspot

WebSockets

advanced

Real-time communication for monitoring and control

RTSP

intermediate

Offline streaming setups, camera integration

Get in Touch

Have a project in mind, want to collaborate, or just say hello? I'd love to hear from you.

Let's Connect

I'm currently open to new opportunities and interesting projects. Whether you're looking for a full-time developer or need help with a specific AI/backend challenge, let's chat!

Location

India

GitHub

@karanparekh11

LinkedIn

/in/karanparekh11

Email

info@karanparekh.com