Cyclistic Case Study
5.5M+ rides analysis with dashboards and recommendations to improve customer conversion and loyalty.
Analytics & ML professional • Assistant Manager at HSBC • ML intern experience
Hi — I’m Advait Chavan. I build production ML systems that deliver measurable impact.
Machine learning engineer with hands‑on experience in computer vision, NLP, and end‑to‑end data pipelines. I design and deploy production‑ready models, interactive dashboards, and automation (e.g., Google Sheets integrations and Plotly Dash) that reduce manual effort, improve decision making, and lower operational risk.
Experienced in model development, data engineering, and cloud deployment — I turn messy data into audit‑ready systems and interactive dashboards.
About Me
I am an analytics and technology professional specializing in machine learning, data science, and advanced Python development. I build predictive models, computer vision systems, and NLP applications that routinely exceed 90% accuracy. I turn data into action by combining system log analysis, automation, and data‑driven insights to optimize workflows and support better decision‑making. I design and deploy end‑to‑end data pipelines, interactive dashboards, and AI‑powered applications that scale within compliance frameworks, improve efficiency, and reduce operational risk. My work focuses on delivering measurable business impact through technical innovation and robust engineering practices.
Skills
Python
React
Tailwind CSS
Machine Learning
Explainable AI
Policy Rule Engines
Netlify
Serverless Functions
Audit Logging
SQL
Dashboarding
Plotly
NLP
Computer Vision
GeoSpatial Analysis
Portfolio Projects
NLP Exam Scoring
Web app with automated scoring for descriptive answers and Google Sheets integration for results management.
Detected: "I Love You" · Confidence: 98%
Sign Language Recognizer (SLR)
Computer vision model to convert gestures to text with real-time webcam support and high confidence detection.
Logistic Regression (best) · Test Acc: 0.8333
Winning Space Race with Data Science
End-to-end analysis of Falcon 9 launches: API + web scraping, EDA, Folium maps, Plotly Dash dashboards, and classification models to predict landing outcomes.
Policy first · Model assisted · Live demo
Online Fraud Detection System
An accessible, auditable fraud scoring system that combines a calibrated ML model with RBI/NPCI and bank policy checks. Single‑transaction scoring, batch CSV processing, and an admin UI to manage merchant categories and bank limits — built for both non‑technical users and data teams.
Hybrid rules + explainable ML for loan decisions
Loan Decision Simulator
Interactive regulator-ready loan decision simulator combining bank rule checks with explainable model scoring and transparent feature-level reasoning.
Experience
Assistant Manager — HSBC
Mumbai, India • April 2025 - Present
Performed system log analysis for fraud investigations, automated ticket workflows, and built complaint trend dashboards to improve SLA adherence and reduce escalations.
Machine Learning Intern — Cloud Counselage
Remote • April 2024 - June 2024
Processed 1,000+ student records; built Random Forest model achieving ~93% accuracy; validated with precision/recall/F1 metrics.