>_

Farhan Sadik

AI & Backend Developer

Building production AI systems -
RAG pipelines, LLM integration, and scalable backends.

View My Work Download CV
Scroll
about me!
Farhan Sadik

I'm Farhan Sadik and I'm an AI & Backend Developer. I specialize in RAG pipelines, LLM integration, and FastAPI backends that scale. I build end-to-end products — from intelligent data systems to containerized microservices. I ship real products, not demos.

SKILLS
Python RAG / PGVector Ollama LLM Integration FastAPI Docker PostgreSQL DeepFace PyTorch JavaScript Linux
EDUCATION
2022–2026
Daffodil International University
BSc Computer Science & Engineering
MEET-FARHAN
Farhan Sadik
SOCIALS
Where I've Worked

Experience

AI Software Developer

Daffodil Computers Limited
Feb 2026 – Aug 2026 On-site · Dhaka, Bangladesh

Engineered and deployed two AI modules within a production HR platform serving internal teams.

Employee Engagement & Retention Analysis

Built a predictive system to evaluate workforce engagement and flag at-risk employees, validated at 97% precision against historical HR outcomes.

Python FastAPI Ollama PostgreSQL HTML/CSS/JS
AI Response Console

Built a RAG-based HR knowledge system using PGVector for semantic retrieval over internal documentation. Validated through an automated pipeline of 24,000+ test queries, achieving 99% accuracy at under 400ms average response time.

Python FastAPI Ollama PostgreSQL PGVector
What I work with

Skills & Technologies

AI & Machine Learning

Python RAG / PGVector Ollama LLM Integration DeepFace NLTK / PyTorch Label Studio

Languages

JavaScript HTML / CSS Java C++ SQL

Backend

FastAPI PostgreSQL REST APIs JWT Auth SSL/TLS

DevOps & Cloud

Docker Docker Compose Linux Git
Selected Work

Projects

Full-Stack Application

PhotoDump.ai

Self-hosted photo-sharing platform for events. Owners create password-protected galleries; guests browse, upload, and download photos. Includes an optional GPU microservice for AI face recognition - guests upload a selfie to instantly find every photo they appear in. Three-tier Docker architecture with cross-host communication and graceful degradation when the GPU service is offline.

Python FastAPI PostgreSQL DeepFace Docker
Case Study → GitHub →
Developer Tool

LocalPilot IDE

Browser-based AI code editor built on Monaco Editor with GitHub Copilot-style ghost-text completions. A FastAPI backend routes requests to a local LLM for inline suggestions, code generation, debugging, and chat. Supports file management and ZIP import/export - all containerized, no data leaves the machine.

Python FastAPI Monaco Editor Docker
Case Study → GitHub →
Voice AI

Fluency.ai

Voice conversation app with a full on-device speech pipeline: Moonshine converts speech to text, Mistral 7B generates a response, Kokoro synthesises the reply as audio. Accessible from any device on the local network via a Gradio HTTPS server - no internet connection, accounts, or API keys required after setup.

Python Moonshine STT Mistral 7B Kokoro TTS Gradio
Case Study → GitHub →
Get In Touch

Let's build something together

Whether you have a project in mind or just want to chat - my inbox is open.

·