Hina Hanan
Machine Learning Engineer with expertise in designing, training, and deploying scalable ML and deep learning models. Strong experience in data preprocessing, feature engineering, model evaluation, and performance optimization using Python, Tensor Flow, Scikit-learn, and SQL.
Proficient in building REST APIs and production-ready ML systems with FastAPI and Docker, and deploying solutions on cloud platforms. Experienced in implementing end-to-end ML pipelines, model monitoring, and AI-driven systems to deliver robust, scalable, and business-impacting solutions.
Experience
Growth Advisor
KPMG
Remote
Mentored students and working professionals in SQL, Power BI, Data Analysis, Business Analysis, and Advanced Excel, delivering structured, industry-oriented training. Designed hands-on learning modules and real-world case studies to strengthen practical understanding of data-driven decision making.
Guided learners in building end-to-end data projects including dashboards, reports, and analytical models. Improved learner outcomes by focusing on problem-solving, data interpretation, and business insights generation.
Data Analyst
Bizbrand Business Consultancy
Kerala
Conducted in-depth data analysis for restaurant businesses across Kerala and international markets, focusing on performance optimization and growth strategies. Performed menu analysis, location analysis, sales analysis, and pricing strategy evaluation to identify revenue opportunities and cost efficiencies.
Built interactive dashboards and reports to track KPIs, sales trends, and customer behavior using data visualization tools. Translated complex data into actionable insights and effectively communicated findings to stakeholders for informed decision-making.
Education
Brototype
Certificate
Indian Institute of technology , Madras
BS Data Science and Programming
Skills
Main Projects
Safe Fi-De Fi Risk assessment Agent
An autonomous De Fi risk assessment agent that monitors protocols and provides actionable insights in real-time. Anomaly Detection: Implemented for De Fi protocols using Autoencoders to identify unusual activity in real-time. Risk Scoring: Computed using LightGBM model, selecting the best-performing model for each individual task.
Alerts: Automated email notifications via SMTP for users subscribing to threshold-based risk scores. Deployment: Dockerized the system and deployed on Google Cloud Platform for scalable, continuous monitoring.
Attend Sys- Face Authentication Attendance System
A secure face authentication attendance system with real-time capture, admin dashboard, and advanced spoof prevention. Backend: FastAPI (Python 3) with Gunicorn + Uvicorn, PostgreSQL using SQLAlchemy ORM. Face Recognition & Processing: dlib-based face_recognition library with 128-dimensional encodings, OpenCV (cv 2) for preprocessing and face detection.
Frontend: React 18 with responsive UI, real-time webcam capture, image upload, and data visualization. Infrastructure & Deployment: Dockerized deployment on GCP, Cloud SQL for PostgreSQL and Cloud Storage for static files
AI News Weekly Digest
A personalized AI-powered weekly news digest aggregating multiple sources with relevance ranking and structured delivery. Implemented multi-source ingestion (RSS, OpenAI News, Anthropic feeds) with transcript fetching and Markdown conversion; LLM outputs validated as Pydantic-structured JSON. Personalized digests per user profile using curator agents to assign relevance scores, ranking, and reasoning.
Durable pipeline with Postgres database to store source items and avoid duplicates; supports multi-recipient scheduling. Generates email-ready Markdown and HTML, delivering clean digests via Gmail SMTP.
Additional Projects
Face-detection-and-recognition-AI
Deep learning system for real-time face detection and recognition via web interface. Implemented full pipeline using Deep Face, Tensor Flow, OpenCV, dlib, Pandas, Num Py, and Streamlit for interactive recognition.
Customer Churn Prediction Dashboard
Streamlit app predicting customer churn with ROI insights and interactive business KPIs. Implemented churn prediction using Logistic Regression (ROC-AUC 0.85) with SHAP explanations; visualized KPIs and retention ROI using Streamlit and Plotly.
Doc-Bot
Production-ready RAG chatbot enabling accurate, context-aware document question answering. Implemented retrieval pipeline using FAISS vector database and Lang Chain with FastAPI backend to generate LLM-powered responses from relevant documents.
Content-AI
Multi-agent generative AI system for producing SEO-optimized blog posts, social captions, and email copy. Built multi-agent pipeline with FastAPI, Streamlit, Google Gemini LLMs, FAISS for RAG review, and PostgreSQL; automated SEO research, content generation, and brand guideline compliance.
Disaster Reflex Trainer
On-chain game simulating disaster scenario reflex testing, presented at Base Hyperthon 2025 National Finals. Developed on-chain interactive game with smart contracts and real-time reflex assessment; recognized at national-level hackathon.
Certifications
Data Science Foundation
IIT Madras
Microsoft Azure AI Essentials Professional Certificate
Microsoft & [PROT 12]
Generative AI: Introduction to Large Language Models
[PROT 13]