Abdulla Jubair
Passionate AI/ML engineer who has built 70+ projects spanning Generative AI, RAG systems, MLOps pipelines, cloud data engineering, and full-stack ML applications. Based in Kerala, India 🇮🇳
Who I Am
Building intelligent systems at the intersection of data science, AI, and cloud engineering.
Turning Data into Intelligence
I'm Abdulla Jubair, a Data Scientist and AI Engineer from Kerala, India. Over the past two years at Brototype, I've immersed myself in the cutting edge of AI — building everything from production-grade RAG pipelines and multi-agent LLM systems to cloud-native ETL architectures and real-time ML deployments.
With 70+ projects across Generative AI, Machine Learning, Data Engineering, BI dashboards, and MLOps, I don't just learn concepts — I build real systems that solve real problems. My capstone, MarketMind AI, demonstrates this: a full-stack stock intelligence platform deployed on Kubernetes with CI/CD, RAG, and multi-agent AI.
I'm passionate about bridging the gap between raw data and business value, whether that's through a production LLM API, an automated Airflow pipeline, or an interactive Power BI dashboard.
Higher Secondary Education
Completed Science stream with an exceptional 97% score, laying a strong foundation in Mathematics and Analytical thinking.
Professional Data Science Training
Intensive hands-on curriculum covering ML, Deep Learning, NLP, Data Engineering, MLOps, LLMs, RAG, Agentic AI, and Cloud technologies.
Python for Data Science Certification
16+ hours of hands-on Python, EDA, data cleaning, statistical plotting, and preprocessing using Pandas, NumPy, Matplotlib, and Seaborn.
My Expertise
A comprehensive toolkit spanning AI/ML, data engineering, cloud infrastructure, and full-stack development.
AI / LLMs
Machine Learning
Data Engineering
Cloud / MLOps
BI & Analytics
Backend / Dev
Flagship Projects
Highlights from 70+ projects — production-ready systems spanning AI, data engineering, and cloud.
MarketMind AI
Stock Market Intelligence Platform
Capstone project — a full-stack, production-grade stock intelligence platform with multi-agent LangGraph reporting, RAG chatbot backed by FAISS and Gemini API, and FinBERT PyTorch sentiment analysis engine. GRU deep learning model for 30-day price forecasting deployed on Kubernetes.
- ✦Multi-agent LangGraph reporting
- ✦RAG + FAISS + Gemini API
- ✦FinBERT news sentiment
- ✦GRU 30-day forecasting
Newslyzer
RAG-based News Research Assistant
AI-powered news research tool that accepts multiple article URLs, chunks content using MiniLM Sentence-Transformers, builds a FAISS semantic index, and answers questions via LangChain RetrievalQA with Mistral-7B. Provides grounded answers with source citations.
- ✦Semantic vector search
- ✦FAISS local index
- ✦LangChain RetrievalQA
- ✦Source citation answers
FlickFinder
NLP-Powered Movie Recommendation System
Content-based movie recommender using NLP and cosine similarity. spaCy lemmatization processes cast, crew, genres, and keywords into dense tag fields. CountVectorizer builds feature matrices for cosine similarity matching. TMDB API integration for dynamic poster and rating display.
- ✦spaCy lemmatization
- ✦CountVectorizer features
- ✦Cosine similarity
- ✦TMDB API integration
Fake News Detector
NLP Classification with GloVe Embeddings
NLP classifier that detects misinformation using spaCy's large GloVe embeddings (en_core_web_lg). Averages 300-dimensional token vectors while filtering stopwords. KNN classifier achieves 99% accuracy. Deployed as a Flask web app for real-time article verification.
- ✦300D GloVe embeddings
- ✦99% accuracy KNN
- ✦Flask deployment
- ✦Real-time article check
Restaurant Name & Dish Planner
LLM-Powered Generative AI App
AI-driven restaurant concept generator using LangChain PromptTemplates and Mistral-7B via Together AI. Multi-step LLM chain links name generation to menu creation. Clean Streamlit UI with sidebar cuisine selectors and formatted dish menus.
- ✦LangChain prompt chains
- ✦Mistral-7B generation
- ✦Multi-step LLM pipeline
- ✦Streamlit UI
IPL Score Predictor
ML Regression with Streamlit App
Machine learning app predicting final IPL match scores from live ball-by-ball data. Evaluated Linear Regression, SVR, KNN, Random Forest, and XGBoost — selecting Random Forest for 93.07% test accuracy. One-Hot Encoding for teams, input validation, and real-time Streamlit predictions.
- ✦93.07% test accuracy
- ✦Random Forest regressor
- ✦One-Hot Encoding
- ✦Streamlit web app
70+ Projects
Browse my complete portfolio organized by technology domain.
SalaryPrediction – AWS Lambda
Docker containerized ML model deployed to AWS Lambda with API Gateway endpoint.
SalaryPrediction – SageMaker
Real-time ML endpoint on Amazon SageMaker with custom inference script and boto3 SDK.
SalaryPrediction – FastAPI + Kubernetes + Helm
Production ML deployment using FastAPI, Docker, Kubernetes, and Helm charts.
SalaryPrediction – FastAPI + Docker
Linear regression model served as REST API with FastAPI and Docker containerization.
Real-Time Prediction – Kafka
Real-time salary prediction streaming experience data through Apache Kafka.
Titanic – FastAPI + Docker
Titanic survival prediction model deployed as REST API with FastAPI and Docker.
MLflow + Evidently
MLOps implementation for experiment tracking, model management, and drift monitoring.
GCP Employee ETL Pipeline
Automated ETL with Cloud Composer (Airflow), Cloud Data Fusion, and BigQuery.
Sales Order ELT – GCP
Data pipeline with Airflow, GCS, BigQuery, and Looker Studio for sales analytics.
Batsman Rankings GCP Pipeline
Real-time ICC cricket batsman rankings via Cricbuzz API → GCS → BigQuery → Looker Studio.
Sales Data GCP Pipeline
Flask web app uploads country-wise sales data to GCS, auto-loads to BigQuery via Cloud Functions.
Sales Analysis – PySpark
Customer insights, product performance, and sales trends analysis using PySpark and EDA.
IPL Score Predictor
ML app predicting final IPL scores using ball-by-ball data with Random Forest (93.07% accuracy).
Bengaluru House Price Predictor
ML web app for house price prediction using linear regression with Flask frontend.
Apple Stock – ARIMA
Time series forecasting of Apple stock closing prices using ARIMA model with EDA.
Fake News Detector
NLP classifier using spaCy GloVe embeddings and KNN achieving 99% accuracy, served via Flask.
LoRA Implementation
Fine-tuning GPT-2 with LoRA adapters demonstrating parameter-efficient training.
Restaurant Name Planner – LLM
AI-driven restaurant name and menu generator using LangChain and Mistral-7B via Together AI.
Netflix Insights Dashboard
Interactive Power BI dashboard with 5,500+ titles analyzing ratings, genres, and global trends.
Sales Insights Dashboard
Power BI dashboard on ₹985M revenue data with DAX measures and profit analysis.
Hospitality Report
Hotel industry KPI dashboard: revenue, occupancy rates, and platform performance analytics.
Airbnb EDA & Report
Pricing trends, room types, and host ratings analysis with Power BI dashboard.
Data Science Job Salary Dashboard
Tableau dashboard visualizing global data science salaries by role, experience, and location.
Zomato Rating Dashboard
Tableau dashboard with Tableau Prep for restaurant ratings, city, and country insights.
Personal Finance Dashboard
Power BI dashboard tracking income, expenses, savings, and net worth trends.
Global Insights Dashboard
Dash and Plotly dashboard for population, GDP, and life expectancy with choropleth maps.
Bookshelf FastAPI
Secure async FastAPI with JWT auth, rate limiting, and full CRUD for book management.
Bookstore Microservices
FastAPI microservices with Prometheus monitoring, structured logging, and PostgreSQL.
Dockerized Bookstore
Fully Dockerized microservices with User, Book, and Order services + Prometheus monitoring.
Kubernetes Bookstore
Bookstore microservices reimplemented on Kubernetes with port-forwarding and Secrets management.
StockWise Inventory System
Full-stack inventory management system with FastAPI, PostgreSQL, and vanilla JS frontend.
E-Commerce SQL Project
SQL analysis on structured e-commerce database to extract customer behavior and sales insights.
RollerCoaster EDA
EDA project on roller coaster data — speed, height, and location trends with Pandas/Matplotlib.
Bengaluru House EDA
Data preprocessing and feature engineering for house price prediction with linear regression.
PySpark Complete Guide
Hands-on guide to PySpark: DataFrames, EDA, ML with Spark MLlib, and streaming.
GitHub CI/CD Calculator
Python calculator project showcasing Git workflows, pytest testing, Pylint, and GitHub Actions.
Let's Connect
Open to Data Science, ML Engineering, GenAI Developer, and AI Engineer roles. Let's build something great together.
Ready to make an impact
I'm actively seeking roles in Data Science, Machine Learning, Generative AI, and AI Engineering. With 70+ real-world projects, I'm ready to contribute from day one.
Ready for the Next Challenge
Whether it's building a RAG pipeline, deploying ML models to production, orchestrating data workflows on cloud, or crafting BI dashboards — I bring hands-on expertise and relentless curiosity to every project.