Tamil Selvan K — AI Developer & Research Portfolio

available for research & engineering roles | Tirupur, TN

Tamil Selvan K

AI Developer · Edge AI · LLM Systems · Computer Vision

Building production-grade AI systems from hardware-level edge pipelines to LLM-driven conversational agents. Focused on measurable outcomes: detection accuracy, deployment scale, and real user impact.

% PCB defect accuracy

% manual inspection reduction

+ users served

production AI products

View Technical Work Get in Touch

§ 01 — professional record

Work Experience

Leadtap

AI Developer

Jan 2026 — Present

Developing speech-to-text, text-to-speech, and generative AI models in production. Currently engineering phone call AI systems that integrate real-time voice pipelines with LLM-driven dialogue management.

KrishTec

AI Developer

Feb 2025 — Dec 2025

Developed a PCB defect detection system achieving 98% accuracy, reducing manual inspection workload by 60%. Owned the full AI model lifecycle — data curation, architecture selection, training pipelines, evaluation protocols, and production deployment.

LuzyTech

Founder & AI Product Engineer

Sep 2022 — Dec 2024

Founded and led a two-product AI startup. Built and shipped Jon-the-Interviewer and YouTube Summarizer from zero to production, collectively serving 800+ users. Handled end-to-end engineering: model selection, API integration, frontend, and deployment infrastructure.

§ 02 — technical artifacts

Selected Projects

ESP-Vision-Pro + AI Rover

Low-latency Edge AI pipeline built on ESP32-CAM for real-time object detection at the hardware level. Integrated AI-triggered rover behavior with gesture-based control UI. Demonstrates end-to-end embedded ML: sensor → inference → actuation.

Edge AI ESP32-CAM Object Detection Embedded ML Robotics

Jon-the-Interviewer

AI-driven mock interview platform with real-time conversational flow powered by LLMs. Integrates speech recognition for live audio input and computer vision models for candidate analysis. Deployed to production with 800+ active users.

LLM Speech Recognition Computer Vision Conversational AI Production

PCB Defect Detection System

Computer vision pipeline for automated quality inspection on printed circuit boards. Achieved 98% detection accuracy with end-to-end ownership of data labeling, model training, and deployment to factory floor. Reduced manual inspection time by 60%.

Computer Vision Quality Inspection Industrial AI Model Deployment

Phone Call AI System

Real-time voice AI pipeline combining STT, TTS, and generative LLM dialogue for automated phone interactions. Focuses on low-latency streaming, turn-taking logic, and natural conversation flow under production constraints.

STT / TTS Generative AI Voice Pipeline Real-time

§ 03 — technical inventory

Skills & Capabilities

AI / Machine Learning

Computer Vision95%

LLM Integration90%

Model Training & Eval88%

Edge AI / Embedded ML82%

Voice & Language

Speech-to-Text (STT)88%

Text-to-Speech (TTS)85%

Conversational Design80%

NLP Pipelines78%

Engineering & Deployment

Model Deployment87%

Data Pipeline Design83%

API Integration85%

Product Engineering80%

Python PyTorch TensorFlow OpenCV Whisper Hugging Face LangChain FastAPI ESP32 Docker Linux Git REST APIs YOLO

§ 04 — academic record

Education

Jai Shriram Engineering College

B.Tech — Artificial Intelligence & Data Science

2021 — 2025

Govt. HSS Perumanallur

Higher Secondary Education

2019 — 2021

§ 05 — contact

Get in Touch

Location

Tirupur, Tamil Nadu, India

Domain

AI Development · Edge Systems · LLM Engineering

Status

Open to opportunities

Made with Resume2Web