Prasanna Patil — Resume

AI Engineer — Zuneko Labs by Emdee Digitronics

June 2026 - Present

• Own end-to-end development of production AI and backend systems: activity-based time tracking and billing, healthcare workflows, conversational/voice agents, document automation, and computer vision.
• Build Python APIs (FastAPI/Django), AI services, automation pipelines, and third-party integrations across databases, messaging platforms, cloud services, and real-time processing.
• Develop and maintain activity-based client billing, WhatsApp document workflows, AI voice agents, and ANPR/CCTV analytics, with a focus on reliability and deployment.
• Hardening internship-era pipelines into monitored, recoverable production services.

Computer Vision Intern — Zuneko Labs by Emdee Digitronics

March 2026 - June 2026

• Built a real-time person detection and tracking pipeline on CCTV feeds using frame-by-frame detection and multi-frame tracking.
• Built a WhatsApp-based document ingestion system using the WhatsApp Business API to pull media from group chats and upload it to Google Drive.
• Wrote a recovery mechanism to backfill missed files after downtime or network failures.

AI/ML Intern (Backend-Focused) — Shivohini TechAI LLP

Sept 2025 - March 2026

• Built backend for conversational AI agents using LLaMA and OpenAI APIs. Focused on keeping context across turns without blowing token budgets.
• Built WhatsApp automation for clients — webhook ingestion, message routing, template management via Meta Business API.
• Moved message processing, AI inference, and DB writes off the request thread with Redis and Celery.
• Migrated client databases from MySQL to Supabase/PostgreSQL mid-project. Kept the AI agents running throughout.
• Talked directly to clients to figure out what they actually needed, not just what they asked for.

Finalist - Backend & CV — Smart India Hackathon (Government of India)

Sept 2023

• Built a CCTV analytics backend for crime detection and crowd monitoring. Presented to a government panel.
• CV pipeline in OpenCV: facial recognition, crowd counting, anomaly detection.
• APIs + background workers for real-time video analysis and alert dispatch.
• Twilio and WhatsApp integration for instant incident notifications.

Speech Emotion Recognition (SER)

Speech Emotion Recognition system using CNN, LSTM, and CLSTM. Trained on CREMA-D, RAVDESS, TESS, and SAVEE — 93.27% accuracy. Real-time inference pipeline for live audio with a Django REST backend and React frontend, aimed at customer service and healthcare use cases.

Case study · Source code

CCTV Analytics for Crime & Crowd Management

End-to-end surveillance platform for public safety. Crowd density estimation, crime detection, facial recognition. MongoDB for event persistence, Twilio and WhatsApp for real-time alerts. Presented at Smart India Hackathon 2023.

Case study · Source code

Agentic Honey-Pot — AI Counter-Scam Platform

Built at GUVI AI Impact Buildathon 2026. LLM-driven personas engage live scammers in real time to stall them and extract threat intelligence. FastAPI orchestration with persona/strategy switching and semantic scam-probability scoring; React 19 + Vite dashboard with API-key auth and rate limiting.

Source code

WhatsApp Automation & AI Agent Platform

Backend for automated WhatsApp workflows — webhook ingestion, user state tracking, async AI responses, conversation history. Redis + Celery for reliability. Built for a client at Shivohini TechAI.