About Experience Education Skills Publications Certifications Open Source Projects Contact
Impact-Driven Strategist

Hi, I'm
Venkatkumar Rajan

Generative AI Engineer with 3.5+ years building production-grade AI systems across multimodal applications, Kaggle Master & author of llmevalkit, passionate about advancing the frontiers of AI through innovative research that transforms data into actionable insights and intelligent systems.

Contact Me Resume GitHub
3.5+
Years Exp.
3
Publications
15+
Projects
Top 1%
Kaggle
Venkatkumar Rajan
About

Who I Am

I'm a Generative AI Engineer passionate about building intelligent solutions that blend research, real-world deployment, and innovation. My work spans Generative AI, Computer Vision, NLP, and end-to-end AI pipelines — from designing algorithms to deploying production-ready systems.

With experience across healthcare, manufacturing, and automotive domains, I've built real-time RAG pipelines, 3D reconstruction systems using NeRF & photogrammetry, document processing platforms on Azure AI, and AI-powered automation tools. I hold an M.Tech in AI from BITS Pilani and am a Kaggle Master (Competitions Expert top ~0.4%, Notebooks Master top ~1.3%).

I enjoy bridging the gap between deep research and applied business impact — with a strong focus on responsible, secure AI systems and governance.

Location
Madurai, Tamil Nadu, India
Email
venkatkumarr.vk99@gmail.com
Kaggle
Master — Top ~0.4% globally
Education
M.Tech AI — BITS Pilani
Status
Currently focused on LLM Evaluation and 3D Computer Vision research
Experience

Work History

Generative AI Engineer | Business Analyst
EXL Service (India) Pvt. Ltd.
Dec 2023 – Present  ·  Noida, India (Remote)
AI-Powered Credential Extraction System
Unified Text RAG + Multimodal RAG · Azure GPT-5 · AI Search · Doc Intelligence
Architected end-to-end unified dual-mode RAG pipeline extracting 46+ structured fields from unstructured docs. Confidence-based routing + human-in-loop review. 70% accuracy · 30% manual effort reduction.
Unified GenAI Document Intelligence & Automation System
Azure · NLP · Computer Vision · Config-driven
Owned end-to-end production. New document type = config entry + prompt update, zero redevelopment. >95% extraction confidence · 40–50% manual effort reduction.
AI-Based Document Classification System
Unified · Azure · >97% accuracy
15+ document categories. New categories via reference document — no engineering changes.
Custom AI Agent Builder
Databricks · Unity Catalog · RAG
No-code agent builder for denial management analytics. Unity Catalog tables + PDF knowledge bases. Governed Databricks App.
Internal AI Productivity Tools + 10+ GenAI PoCs
Excel AI · SQL Chatbot · Power App · Doc Intelligence
Suite of internal tools reducing manual effort. 10+ PoC presentations to prospective clients.
Machine Learning Engineer
AugRay
Sept 2022 – Oct 2023  ·  Chennai, Tamil Nadu (Onsite)
Real-Time Ball Fault Detection
Edge AI · MobileNetV2 · NVIDIA Jetson AGX
90%+ accuracy · QA error rate –40% · throughput 3×. Lightweight real-time defect detection, edge-optimised.
Automated 3D Reconstruction
Generative AI · Photogrammetry · NeRF
End-to-end pipeline: photogrammetry + NeRF. 95% accuracy for non-reflective, ~50% for reflective objects.
Wall & Floor Segmentation
SegFormer-B0 · segformer-b0-finetuned-ade-512-512
Fine-tuned SegFormer-B0. 80% accuracy · designer effort –20+ hrs/month. Automated colour preview generation.
Foot Detection & Virtual Shoe Try-On
Computer Vision · AR · 10,000+ annotations
79% placement accuracy · customer returns –5% in pilot trials.
Generative AI PoCs
Stable Diffusion · Face Texture · Chatbot
Automated flyer generation, face texture synthesis, site-based conversational chatbot.
Education

Academic Background

M.Tech. in Artificial Intelligence & Machine Learning
Birla Institute of Technology and Science (BITS), Pilani
2023 – 2025  ·  Rajasthan, India
Coursework: Deep Learning, Reinforcement Learning, Probabilistic Generative Models, Graph Neural Networks, Video Analytics, Social Media Analytics, Distributed ML
B.E. in Electronics & Communication Engineering
SACS MAVMM Engineering College, Anna University
2016 – 2020  ·  Madurai, Tamil Nadu
Coursework: Data Structures, Algorithms, Digital Signal Processing, Digital Image Processing, Embedded Systems, Electronics
Skills

Technical Stack

Programming Languages
PythonC++
LLM Systems & AI Engineering
RAGGraphRAGAI AgentsLLM EvaluationPrompt EngineeringEnd-to-End OrchestrationMonitoring
LLM Frameworks
LangChainLlamaIndexLangGraphMCPGroqOllamaDocling
ML / DL Frameworks
PyTorchTensorFlowOpenCVscikit-learnKornialibrosaMediapipePySparkNetworkXGymProphet
Cloud & Platforms
Azure OpenAIAzure AI SearchDoc IntelligenceAzure BlobKey VaultAzure DevOpsAWSDatabricks
MLOps & LLMOps
DockerKubernetesKubeflowMLflowKafkaCI/CD
Edge AI
NVIDIA Jetson AGX XavierTensorRTRaspberryPi
Web & Backend
FastAPIFlaskDjangoREST APIsStreamlitGradioReactJS
Database & Tools
MongoDBMySQLPower BIPower AutomatePower AppsPytest
Research

Publications

arXiv arXiv · March 2025
A Generative Approach to High Fidelity 3D Reconstruction from Text Data
Text-to-3D pipeline: Stable Diffusion → RL enhancement → Stable Delight → volumetric 3D reconstruction for AR/VR applications.
Click to read →
IEEE IEEE · Feb 2024
Advancing Audio Fingerprinting Accuracy with AI and ML
Addressing background noise and distortion challenges in audio fingerprinting using advanced AI and ML approaches.
Click to read →
IRJIET IRJIET · March 2020
PCB Layout using CNC Machine Controlling with Wireless Communication
Implementation of automated PCB layout using CNC machine with wireless control systems.
Click to read →
Credentials

Certifications

TensorFlow Developer
Google · Oct 2023 – Oct 2026
View Credential →
Nvidia Jetson AI Specialist
Nvidia Deep Learning Institute · March 2023
View Credential →
Azure AI Fundamentals
Microsoft · April 2024
View Credential →
Self Driving Car Engineer
Udacity Nanodegree · January 2025
View Credential →
Reinforcement Learning Specialization
University of Alberta · Oct 2024
View Credential →
Generative AI for Software Development
DeepLearning.ai · Oct 2024
View Credential →
Open Source

llmevalkit

llmevalkit
PyPI v5.0 Active

Universal Python library for evaluating and benchmarking LLM systems across 61 metrics in 10 modules covering Quality, Hallucination Detection, Compliance, Document AI, Governance, Security, Multimodal Evaluation, AI Content Detection, Observability, and Anomaly Detection. Works fully offline or with any LLM provider, supports RAG pipelines, AI agents, document extraction systems, and requires no ground-truth labels.

v5.0 Highlights: Added AI Content Detection, Observability, and Anomaly Detection modules, expanded hallucination detection to 12 metrics, introduced auto-logging, drift analysis, prompt injection detection, HIPAA/GDPR/DPDP compliance checks, multimodal evaluation, and LLM-as-judge support — all in one unified framework.

61 evaluation metrics
Offline + LLM-as-judge modes
35+ presets & 8 providers
RAG, Agents & Multimodal ready
PyPI GitHub
pip install llmevalkit

from llmevalkit import Evaluator

# Offline evaluation (free, no API)
evaluator = Evaluator( provider="none", preset="hallucination")
result = evaluator.evaluate(question, answer, context)
print( result.summary())

# LLM-as-judge evaluation
evaluator = Evaluator( provider="openai", model="gpt-4o-mini", preset="production")
result = evaluator.evaluate(question, answer, context)
print( result.summary())
Projects

Featured Projects

Click any card to explore. For the full portfolio visit github.com/VK-Ant.

3D
Generative AINeRF3D Vision
Text-to-3D Reconstruction Pipeline
Text → Stable Diffusion → RL enhancement → Stable Delight → 3D reconstruction. arXiv published.
CV
Computer VisionDeep Learning
Comprehensive Computer Vision Portfolio
Object detection, multi-object tracking, and real-time video analysis solutions.
Attendance
Edge AIJetson AGX
Face-Based Attendance System
Facial recognition attendance system on NVIDIA Jetson AGX Xavier for field employees.
RAG
RAG/CAGKnowledge GraphMultimodal
GenAI RAG/CAG + Knowledge Graph & Multimodal Chatbot
PDF/CSV/Vision drag-drop chatbot with multilingual text + audio. Security and drone domains.
Vehicle
Video AnalyticsKITTI
Advanced Video Analytics for Vehicle Behavior
Vehicle behaviour analysis using KITTI dataset with advanced CV and ML.
Recruitment
Real-time CVEmotion AI
AI-Powered Real-Time Video Recruitment Tool
Eye contact, emotion, age, gender recognition for data-driven AI recruitment.
Contact

Get In Touch

Always open to new opportunities, collaborations, and meaningful conversations about AI.

Email
venkatkumarr.vk99@gmail.com