Hi, I'm

Harigovind P

AIEngineer|LLMSystems&Pipelines

Building production AI systems — RAG pipelines, LLM inference layers, and reliable ML infrastructure — with strong backend engineering foundations.

Download Resume

About Me

I'm a final-year B.Tech student in Artificial Intelligence and Machine Learning, focused on building production-grade AI systems — from LLM pipelines to reliable retrieval infrastructure.

My work sits at the intersection of LLM system design, RAG engineering, and backend development — building systems with observability, self-healing loops, and measurable reliability baked in from the start.

Looking For

AI Engineer / LLM Systems Roles
RAG & ML Pipeline Engineering
Backend & ML Integration Roles
Production AI System Development
Research-to-deployment ML projects

LLM Systems & RAG Pipelines

Designing self-healing retrieval pipelines with confidence scoring, reranking, and hallucination detection.

ML Model Development

Computer vision and NLP systems — PyTorch, OpenCV, and scikit-learn for robust perception and classification.

Backend + ML Integration

FastAPI services, structured logging, and deployable ML — bridging model development to production systems.

Projects

Production AI systems, ML pipelines, and full-stack applications — built end-to-end with reliability and observability in mind.

Flagship System

Featured SystemProduction-grade

Adaptive Knowledge Intelligence System (AKIS-v2)

RAGLLMFAISSSelf-HealingObservability

Architected a production-grade self-healing RAG pipeline with modular ingestion, hybrid FAISS+BM25 retrieval, cross-encoder reranking, and LLM generation. Deployed on CPU with full observability via structured JSON logging.

Implemented an automated confidence-scoring and query-rewriting loop that detects low retrieval quality, rewrites queries via the LLM, and retries before falling back to an uncertainty-aware response — eliminating hallucination risk without human intervention.

Built a semantic hallucination detection layer that splits generated answers into atomic claims and verifies each against retrieved context using cosine similarity, emitting structured {grounded, confidence, explanation, claims[]} JSON for downstream auditability.

Exposed as a FastAPI service with /query, /ingest, and /metrics endpoints, enabling background PDF re-indexing, live query serving, and lightweight performance monitoring.

RAGFAISSBM25FastAPILLMPythonCosine Similarity

Featured Projects

Featured

Smart Water Leak Detection System

End-to-end IoT pipeline for real-time anomaly detection across water infrastructure. ESP32-driven sensor mesh feeds an LSTM model for predictive leak detection, surfaced via a Flask monitoring dashboard.

ESP32IoTLSTMPythonFlask
Featured

AgriSkill – Expert Matching Platform

Web platform connecting farmers with domain experts using intelligent matching. TF-IDF + k-NN recommendation engine surfaces relevant expertise, reducing search friction in agricultural advisory workflows.

DjangoPythonTF-IDFk-NN
Featured

Motion Detection & Alert System

Automated surveillance pipeline that detects motion from live video feeds and triggers structured email alerts with image capture — reducing manual monitoring overhead significantly.

OpenCVPythonSMTP

Other Notable Projects

Real-Time Violence Detection System

CNN-LSTM architecture for detecting violent activity in multi-person video streams. Spatial feature extraction via CNN combined with LSTM temporal modeling for frame-sequence classification.

PyTorchOpenCVCNNLSTM

Cyberbullying Tweets Classifier

NLP classification pipeline using TF-IDF vectorization and Random Forest for cyberbullying detection across tweet datasets. Demonstrates applied NLP for content moderation use cases.

PythonNLPScikit-learnTF-IDF
AI-assisted

Thengha Launchpad

Full-stack application built with Loveable AI platform — demonstrating rapid AI-assisted development workflows and production-ready deployment patterns.

ReactTailwind

Research & Publication

IEEE Conference

Real-Time Water Leak and Usage Monitoring for Smart Apartments using IoT Flow Sensors and LSTM-based Leak Detection

Presented at IEEE International Conference — IDCIoT 2026

Skills

AI/ML system engineering with strong backend, retrieval, and LLM infrastructure experience.

Programming Languages

PythonAdvanced
JavaIntermediate
SQLIntermediate
HTMLIntermediate
CSSIntermediate
CBasic

AI / LLM Systems

RAG Pipelines
LLM Integration
Vector Search (FAISS)
BM25 Retrieval
Hallucination Detection
Prompt Engineering

ML & Deep Learning

PyTorch
Scikit-learn
OpenCV
NumPy
Pandas
NLP
Computer Vision

Backend & APIs

FastAPI
Flask
Django
REST APIs
Structured Logging
JSON Observability

Tools & Platforms

Git
GitHub
Linux (Ubuntu)
Jupyter Notebook
Google Colab
Docker (basics)

Databases

MySQL
SQLite
Vector Stores

Certifications

Credentials and continuous learning in cloud, AI, and software engineering.

AWS Academy Graduate – Cloud Foundations

Understanding of cloud computing concepts, AWS services, and deployment fundamentals.

View Certificate

Mastering Generative AI and ChatGPT

Covers prompt engineering, generative AI concepts, and real-world applications of large language models.

View Certificate

Get in Touch

Let's talk about everything!

Open to collaborations, AI/ML opportunities, and backend development roles. Feel free to reach out — I'll get back to you as soon as possible.

Email

Location

Kerala, India

Based in India • Available Worldwide