Omkar
Bandikatte
Building RAG pipelines, AI agents, and full-stack products that turn complex workflows into reliable software.
About Me_
I build AI systems that ship: RAG pipelines, tool-calling agents, and the full-stack products around them. As a Computer Science (Data Science) student at DJSCE and former SDE Intern at Periscope Technologies, I have built production automation, data pipelines, and trustworthy AI experiences across e-commerce, recruiting, research, and computer vision.
Skills_
MY TECH STACK
// Languages
// Web Development
// Tools & Frameworks
// AI / ML
Projects_
DocuMind — Multi-Document Intelligence
Trust-aware RAG assistant across PDFs, web pages, YouTube, audio, CSVs, and images. Hybrid Qdrant and Elasticsearch retrieval uses Reciprocal Rank Fusion, citations, credibility scoring, temporal drift detection, and self-verification.
LeadIntel — B2B Lead Intelligence
End-to-end lead scoring platform built from 6,711 publicly sourced company records. Combines conversion prediction, local Hugging Face intent classification, role-based APIs, automated retraining, and ranked account analytics.
Experience_
SDE Intern

- →Led development of Stryder, a multi-client e-commerce analytics platform, building Playwright automation across Blinkit, Zepto, Instamart, BigBasket, Amazon, and Flipkart Minutes.
- →Built FastAPI data pipelines processing 1,000+ records per day, with consolidated reporting in Azure Blob Storage and PostgreSQL and sub-100ms reporting queries.
- →Designed an LLM-assisted self-healing recovery module with cached regex rules and keyword fallbacks, improving pipeline reliability by 35%.
- →Implemented secure multi-client authentication, TOTP/2FA automation, encrypted credential handling, and fixes for concurrent shared-inbox OTP races.
Achievements_

1st Runner Up — Digital Healthcare Periscope Hackathon

2nd Runner Up — Xtract 3.0 Data Science Competition
Top 10 in HackWithMumbai Hackathon
AI Smart City Hackathon (ByteCamp25): Integrated waste, flood, and solar systems in 24 hours
Scroll Hack: Built Dout-SirJi AI chatbot with blockchain-based login
Qualified Technograd Round 2 — Data Science and Machine Learning Challenge
Co-Curricular_

Events Head
Led event planning and execution for DJS Codestars, the coding club of DJ Sanghvi College of Engineering. Organized hackathons, workshops, and competitive programming contests.
Education_
Bachelor of Technology in Computer Science and Engineering (Data Science)
SVKM's Dwarkadas J. Sanghvi College of Engineering
Resume_
Get In Touch_
Interested in working together? Drop me a message.