Agam Singh Saini
Data Scientist & ML Engineer
Master’s candidate in Data Science bridging the gap between Mechatronics Engineering and advanced Machine Learning. Architect of high-speed data pipelines, predictive models, and Sydly—a live transit ecosystem tracking 10,000+ vehicles with 5-second precision.
Technical Architecture
Core Languages
Machine Learning
Deep Learning, Neural Networks, NLP, Computer Vision, TensorFlow, Scikit-Learn, XGBoost, CatBoost, LightGBM
Data & Ecosystem
Pandas, NumPy, PySpark, FastAPI, Streamlit, ROS2, Protocol Buffers, OpenStreetMap
Cloud & DevOps
Git, CI/CD, Docker, Databricks, Azure, Power BI (DAX)
Experience Timeline
Support Worker (Data Context)
Informed Care Services | Sydney
- Managed detailed care plans for 5+ clients, ensuring 100% documentation accuracy.
- Acted as a proactive communicator resolving complex scheduling logistics alongside full-time studies.
Robotics R&D Intern
Combat Vehicle R&D Establishment
- Designed autonomous computer vision pipelines using ROS2, implementing real-time detection improving safety by 40%.
- Simulated 50+ hours of behavior in Gazebo, deriving data-driven prototype improvements.
Mechatronics Intern
Spectrum Technology
- Developed IoT data acquisition systems in LabVIEW to monitor 50+ industrial sensors in <1s.
- Built continuous alert mechanisms achieving 99.9% verifiable uptime.
Academic Foundation
M.S. in Data Science & Innovation
University Of Technology Sydney
B.E. in Mechatronics Engineering
Thapar Institute of Engineering and Tech.
Model Deployments
Sydly-Travel Sydney Seamlessly
Live BuildArchitecture: Geospatial Data & API
Successfully launched a comprehensive transit application. Engineered a high-speed backend consuming live TfNSW GTFS-R feeds, processing and mapping 10,000+ active vehicles with a 5-second telemetry refresh rate.
NYC Taxi Analytics
PipelineArchitecture: Big Data & ML
Engineered a scalable Databricks pipeline to process and analyze 1 Billion taxi records, successfully cleaning over 80M rows of raw data. Developed fare prediction models achieving 87% accuracy.
Crypto Predictor
ModelArchitecture: Time-Series Forecasting
Deployed a live cryptocurrency forecasting web app. Trained ensemble ML models on a decade of historical data, achieving a 79.6% R² score for next-day price predictions.
Weather Insight App
Architecture: Ensemble Learning
Developed a predictive system providing reliable forecasts for rain occurrences using ensemble ML models and the live Open-Meteo API.
DoomBuoy Lib
Architecture: PyPI Open Source
Created and published a Python package designed to significantly accelerate everyday data preprocessing and ML workflows.
Credentials
Specializations
IBM, DataCamp, UTS
NPTEL
ML & Analytics
Online Courses
U-Michigan & IBM
Specializations
UTS BUILD Graduate
Nov 2025
Data Scientist Pro
Jun 2025
IBM Data Science
May 2023
NPTEL Courses
Data Analytics (Python)
Jul 2023
Intro to ML
Jul 2023
Online Courses
Applied Plotting
Jul 2023
SQL for Data Science
May 2023
Initiatives
UTS Peer Networker
Guided 1,000+ new students during orientation.
Media Ambassador, UTS BUILD
Translated complex program missions into engaging stories.
SOUL Program Volunteer
University of Technology Sydney · Feb 2025 – Present · Supported community engagement, equity, and student wellbeing events across campus.
Clubs & Societies
Member of the following societies at the University of Technology Sydney