NOBIL S.
AI & Data Science Student | Aspiring AI/ML Engineer

Building practical solutions with Machine Learning and AI.

I'm a final-year Artificial Intelligence and Data Science student interested in Machine Learning, NLP, and Generative AI. I enjoy building practical applications using Python, exploring AI technologies, and solving real-world problems through software.

View ProjectsGet in Touch
github.com/Nobil-libon
linkedin.com/in/nobil-s
Final-Year B.Tech · Class of 2027
PROFILE SNAPSHOT // NOBIL S.
OPEN TO ROLES
B.Tech Artificial Intelligence & Data Science
J. J. College of Engineering and Technology (2023–2027)
CGPA: 8.05·Tiruchirappalli, India
Primary Areas of Interest
Machine LearningNatural Language ProcessingBERT Fine-TuningRAG SystemsVector EmbeddingsPython Development
Applied Projects Built
Hands-on experience fine-tuning BERT for multi-class emotion classification, developing document ingestion and RAG search with pgvector, and analyzing streaming event anomalies.
Seeking Entry-Level AI/ML & NLP Roles
Learn More ↓
01 // ABOUT ME

AI & Data Science student with a focus on practical applications.

I am a final-year B.Tech student in Artificial Intelligence and Data Science at J. J. College of Engineering and Technology (graduating in 2027 with a CGPA of 8.05).

My primary technical interests are in Machine Learning, Natural Language Processing, and Generative AI. Rather than focusing solely on theoretical models, I enjoy applying these technologies to build working software solutions — including fine-tuning BERT models for text classification and implementing Retrieval-Augmented Generation (RAG) pipelines for semantic document search.

I am eager to begin my career as an entry-level Machine Learning Engineer, AI Engineer, or NLP Engineer where I can continue learning, contribute to real-world projects, and grow alongside an engineering team.

DegreeB.Tech AI & DS
CollegeJJCET, Trichy
Graduation & CGPA2027 · 8.05 CGPA
Target RolesEntry-Level AI/ML

Machine Learning & NLP Foundations

Solid conceptual understanding of supervised learning algorithms, model evaluation (precision, recall, F1-score), cross-validation, and transformer architectures. Practical experience fine-tuning BERT for multi-class classification.

Generative AI & RAG Pipelines

Hands-on work building document ingestion pipelines with PDF extraction, semantic chunking, generating vector embeddings, and querying PostgreSQL with the pgvector extension for cosine similarity search.

Supporting Backend & Software Skills

Proficient in Python, FastAPI for building REST APIs, relational databases (PostgreSQL, SQLite), version control with Git, and clean modular code architecture to turn models into working software.

02 // SELECTED WORK

Featured Projects

Practical applications across Natural Language Processing, Generative AI (RAG), anomaly detection, and full-stack software development.

01 // Generative AI & Platform

DumbWell — Gym Management Platform with RAG Assistant

Multi-Tenant Platform & Document-Based AI Retrieval

Developed a gym management platform and integrated a RAG-based AI assistant for document-based information retrieval.

Key Contributions:
  • Built a document ingestion pipeline with PDF and text extraction.
  • Implemented semantic chunking and vector embeddings using the Gemini API.
  • Integrated semantic retrieval using PostgreSQL pgvector for accurate passage lookup.
  • Implemented role-based access and tenant-level data isolation.
FlutterPythonSupabasePostgreSQLGemini APIGemini Embeddingspgvector
Visit DumbWell.in
RAG Information Retrieval Pipeline
INGESTION
PDF & Text Extraction → Semantic Chunking
EMBEDDINGS & STORAGE
Gemini Embeddings Stored in PostgreSQL pgvector
QUERY & RETRIEVAL
Semantic Similarity Search (Cosine Distance) + Tenant Isolation
BERT Classification Workflow
STEP 1: INPUT TEXT
Raw Sentence Input
STEP 2: PREPROCESSING
WordPiece Tokenization, Padding & Attention Masks
STEP 3: PRETRAINED MODEL
Fine-Tuned BERT Transformer Encoder
STEP 4: OUTPUT PREDICTION
Joy · Sadness · Anger · Fear · Surprise
02 // NLP & Machine Learning

Emotion Detection Using BERT

Transformer-Based Text Classification

Developed a BERT-based NLP model to classify emotions in text, including joy, sadness, anger, fear, and surprise.

Key Contributions:
  • Preprocessed text using BERT tokenization (WordPiece tokenization, truncation, and attention masks).
  • Fine-tuned a pretrained BERT model for multi-class emotion classification across five emotion categories.
  • Implemented inference to predict emotions from custom text inputs.
PythonPyTorchHugging Face TransformersBERTNLP
GitHub Profile
03 // Applied AI & Event Analysis

Streaming Event Anomaly Detector

Event Analysis & Anomaly Detection Workflow

Developed an application to analyze event data and identify unusual patterns using anomaly detection techniques.

Key Contributions:
  • Implemented backend APIs using FastAPI for asynchronous event ingestion.
  • Developed an event analysis and anomaly detection workflow to flag outlier activity.
  • Integrated local LLM capabilities using Ollama (Llama 3.1) for incident diagnostic summaries.
PythonFastAPIReactSQLiteOllamaLlama 3.1
GitHub Repository
Detection Pipeline Overview
FASTAPI INGESTION
Continuous Event Logging & Validation
ANALYSIS ENGINE
Sliding Window Threshold & Anomaly Flagging
LOCAL LLM EXPLANATION
Ollama (Llama 3.1) Incident Root-Cause Summary
04 // FULL-STACK E-COMMERCE PLATFORM (SUPPORTING PROJECT)
Live Website

Sastha Food Products — E-commerce Platform

Built a full-stack e-commerce platform with product management, customer workflows, order processing, and payment integration.

Next.jsReactTypeScriptSupabasePostgreSQLRazorpay
03 // TECHNICAL SKILLS

Technical Skills & Stack

Organized into five core categories reflecting my coursework, hands-on projects, and practical development experience.

Programming & Data Analysis

PythonSQLNumPyPandasMatplotlib

Machine Learning

Scikit-learnSupervised LearningClassificationRegressionKNNRandom ForestSVMCross-ValidationRegularizationFeature ScalingCategorical Encoding

Deep Learning & NLP

PyTorchBERTHugging Face TransformersNLPTokenizationText PreprocessingFine-Tuning

Generative AI

RAGLLMsGemini APIVector EmbeddingsSemantic SearchpgvectorPrompt EngineeringOllama

Backend & Tools

FastAPIREST APIsPostgreSQLMySQLSQLiteSupabaseGitGitHubDocker
04 // CONTACT

Let's connect.

I am actively seeking entry-level Machine Learning Engineer, AI Engineer, and NLP Engineer positions. Whether you have an open role, questions about my projects, or want to discuss applied AI, I would be happy to talk.

Open to entry-level AI/ML roles
GitHubgithub.com/Nobil-libon
LinkedInlinkedin.com/in/nobil-s