Fiaz Bhuiyan
AI/ML Engineer - Geospatial Intelligence & Autonomous Driving
General Motors
The University of Texas at Austin
The University of Texas at Dallas
The Institute of Electrical and Electronics Engineers
About Me
I’m an AI/ML engineer who enjoys working at the intersection of large-scale data systems, geospatial intelligence, and real-world engineering. At General Motors, I focus on advancing SuperCruise mapping technologies by building scalable algorithms and data pipelines that make autonomous driving safer and more reliable.
On the SuperCruise Maps team, I help develop the systems that detect and fix HD map anomalies before they reach customers. My team’s work played a key role in expanding SuperCruise to over 589,000 miles and reducing map-related escalations by 93%, contributing to MotorTrend’s “Best Tech Award.” Most of my day-to-day involves architecting pipelines, writing code, and turning ML ideas into production-ready systems that can handle petabyte-scale geospatial data.
I’m also currently pursuing a Master’s in AI/ML at UT Austin, which gives me the chance to deepen my understanding of the math and theory behind the tools I use at work—everything from probability and statistics to modern LLM and VLM architectures.
Beyond work, I’m passionate about coffee—everything from roasting and brewing to reading about different coffee cultures and, of course, drinking great coffee. I’m also an amateur badminton player who competes in local tournaments, currently wielding an Arcsaber 11 Pro. In my free time, I enjoy spending quality time with family, reading books, and exploring new places and cuisines.
What motivates me most is solving meaningful problems with thoughtful people. I’m drawn to challenges where AI, large data systems, and real-world impact meet, and I enjoy being part of teams that take ideas from concept to production.
M.S. Artificial Intelligence, 2026
The University of Texas at Austin
B.S. Computer Engineering, 2021
The University of Texas at Dallas
IGCSE A-Levels, 2016
Bangladesh International School & College
- Large Language Models (LLMs)
- Vision-Language Models (VLMs)
- Geospatial Intelligence & Mapping
- Autonomous Vehicle Systems
- Distributed ML Systems
- Computer Vision
- Deep Learning Architectures
- Leadership & Innovation
Experience
Leading development of advanced geospatial intelligence systems:
- HD Map Intelligence: Led development of in-house pipeline to detect and remediate geometric anomalies in supplier HD maps using Databricks, PySpark, Sedona, Shapely, and PyProj, expanding SuperCruise coverage to 589,809+ miles for 420,000+ users
- Impact & Recognition: Remediated 1,284 critical locations, achieving 93% reduction in map-related escalations and 3x increase in customer engagements, contributing to MotorTrend’s “Best Tech Award”
- Big Data Infrastructure: Developed fully automated, scalable ETL pipeline to ingest high-volume HSVT vehicle telemetry data from on-premises to Databricks Delta Lake using PySpark, Kubernetes, Azure Synapse, reducing data latency to under 24 hours
- Cost Optimization: Implemented automated data governance and lifecycle management for ADLS and Delta Lake, achieving 600TB storage reduction and $100K annual savings
- ML Innovation: Developed ML models progressing from heuristic algorithms to advanced clustering (HDBSCAN) to infer driving behaviors from HSVT data for optimal vehicle selection
Developing production software for autonomous driving systems:
- Production Software: Developed C/C++ production software for Driver Monitoring System in SuperCruise autonomous driving platform
- Cost Savings: Created automatic after-market hardware detection algorithm for trucks, saving $1.5M annually in technician costs
- System Integration: Collaborated with cross-functional teams to integrate safety-critical software components
Building IoT systems with machine learning integration:
- IoT Development: Built IoT sensor network with LoRa communication for room occupancy prediction
- Full-Stack Solution: Deployed Python MQTT parser on Linux server with SQL database, integrating hardware sensors with deep learning model
- ML Performance: Achieved 90%+ prediction accuracy for occupancy detection system
Responsibilities include:
- Develop custom software to automate data collection using cross-platform APIs
- Synthesize multimodal imagery to create coherent pictures guiding clinical decision making and interventions
- Use Deep Learning to process medical images
Responsibilities include:
- Develop efficient software for the real-time processing of the embedded systems used in prostheses and exoskeletons
- Test and debug existing code to integrate new functionalities like wireless control protocols
Responsibilities include:
- Develop the hardware and software for a telemedicine web-server
Acheivements
Recent & Upcoming Talks
Skills
LLMs, VLMs, Deep Learning
Apache Sedona, Shapely, PyProj
PySpark, Databricks, Distributed Computing
SuperCruise, UltraCruise, HD Mapping
Docker, Kubernetes, CI/CD
Advanced ML Pipelines
IEEE President, Team Management
SPIE Publications, Algorithm Design
Creative Visual Storytelling