// Data Engineer & ML Practitioner

Siddhesh
Chavan

I build scalable data pipelines, ML models, and AI-powered systems that turn raw data into real decisions. MS in Data Science from Pace University — currently engineering data infrastructure at Community Dreams Foundation.

View my work Get in touch
Siddhesh Chavan

NYC skyline · 2024

Open to new opportunities
Python·SQL·Azure·AWS·GCP·FastAPI·PostgreSQL·ETL Pipelines·Machine Learning·LangChain·CrewAI·Docker·BigQuery·Kafka·Tableau·RAG· Python·SQL·Azure·AWS·GCP·FastAPI·PostgreSQL·ETL Pipelines·Machine Learning·LangChain·CrewAI·Docker·BigQuery·Kafka·Tableau·RAG·

01 — About

Data pipelines,
ML models &
AI systems.

I'm a Data Engineer with a background in Electronics & Telecom and a Master's in Data Science. I specialize in building robust ETL/ELT pipelines, designing data warehouses, and applying ML to solve real-world problems.

Currently at Community Dreams Foundation, I work on Azure-based pipelines processing millions of records. I'm passionate about the intersection of data engineering and AI — from classical ML to LLM-powered agents with CrewAI and LangChain.

Pace University — New York, NY
MS in Data Science | GPA: 3.52 / 4.0
May 2025
K.J. Somaiya Institute of Technology — Mumbai
B.Tech in Electronics & Telecommunication | GPA: 8.96 / 10
June 2023

Data & Cloud

  • Python / SQL Expert
  • ETL Pipelines / Data Warehousing Expert
  • Azure / AWS / GCP Advanced
  • PostgreSQL / BigQuery / Redis Advanced
  • Docker / GitHub Actions Advanced

AI & Machine Learning

  • LangChain / LangGraph / RAG Advanced
  • CrewAI / MCP Agents Advanced
  • Random Forest / Regression / KNN Advanced
  • FastAPI / Flask API / REST APIs Advanced
  • Tableau / Looker Studio Intermediate

02 — Projects

001

Patient Health Management System

Healthcare backend system helping clinics, doctors, and patients manage medical records, appointments, and health data efficiently.

  • Reduced manual appointment handling by 50–60% through automated slot generation
  • Minimized scheduling errors by 90% and cut operational costs by 25–35%
  • AI-agent integration using CrewAI and MCP for intelligent scheduling
FastAPICrewAIMCPPostgreSQLPython
↗

002

NYC Trip Data & Business Insights

End-to-end data pipeline for NYC taxi data, transforming 100k+ raw records from GCS into BigQuery analytics and interactive dashboards.

  • Identified 15% increase in peak-hour rides using optimized BigQuery SQL
  • Improved stakeholder reporting speed by 60% via Looker Studio dashboards
  • Automated ingestion and transformation using Mage AI orchestration
Mage AIBigQueryGCPLooker StudioSQL
↗

003

NVidia Stock Price Prediction

Random Forest model for stock price prediction with advanced feature engineering using lagged prices and moving averages.

  • Achieved R² of 0.85, capturing 85% of price variability
  • Improved accuracy by 10% through Close_Lag1/2 and MA5/10/20 features
PythonRandom ForestFeature EngineeringSklearn
↗

004

Big Data Airline Delay Analysis

Large-scale airline performance analysis using AWS infrastructure, predicting flight delays with ensemble ML models.

  • 93% accuracy & 90% precision — Random Forest outperformed all other models
  • Used AWS S3 for storage and EC2 for compute with SQL-based CLI queries
AWS S3AWS EC2Random ForestLogistic RegressionSQL
↗

03 — Experience

Current Sep 2025 — Present
New York, NY

Data Engineer

Community Dreams Foundation

  • Implemented Azure-based data pipeline processing millions of records into data warehouses for analytics & reporting.
  • Improved pipeline performance by 20–30% through optimization and scalability enhancements with comprehensive documentation.
  • Developed well-structured data models that reduced query latency and improved analytical efficiency.
  • Executed end-to-end testing and deployment workflows to ensure high data quality and production readiness.
Internship Jun 2021 — Aug 2021
Mumbai, India

IoT Intern

Verzeo

  • Developed 2 IoT projects integrating sensor data collection and real-time transmission — improved data accuracy by 20%.
  • Processed and analyzed 1,500+ sensor data points using Python for automation tasks with 87% execution accuracy.
  • Built 2 Google Assistant voice-controlled applications to trigger IoT devices based on 10+ predefined conditions.
  • Created IFTTT applets for automating IoT workflows with 5+ data-driven decision triggers.

04 — Contact

Let's build
something.

Open to full-time data engineering & ML roles, interesting projects, and conversations about hard data problems.

✉ Email ◈ GitHub ⬡ LinkedIn ◉ 551-200-1176