Software Engineer | Machine Learning Researcher
Prateek Kumar Goel
I am passionate Machine Learning Researcher and Software Engineer. With expertise in diverse technologies like Python, C++, React, and PyTorch, I excel in creating innovative solutions that drive efficiency and growth.
connectprateek95@gmail.comAbout
- Gainesville, FL (Open to Relocate)
- Linkdin: /pkgprateek
- Github: /pkgprateek
- Medium /prateekkumargoel
- Twitter /prateekkgoel
Contact Me
Education
-
University of Florida
MS, Computer Science (2021-2023) Software Engineering, Advanced Data Structures, Algorithms, Distributed Systems, Machine Learning, Engineering Project Management -
BML Munjal University
BTech, Computer Science (2014-2018) Artificial Intelligence, Computer Networks, Internet of Things, Computer Architecture, Mobile Application Development
Skills
Languages
- HTML/CSS
- JavaScript
- React.js
- Node.js
- GitHub
- MongoDB
- VS Code
-
Machine Learning Reseacher
University of Florida (Jul 23 - Present)- ● Achieved a 15% reduction in resource utilization costs through custom transformer architecture, integrating autoregression for precise adaptation to seasonal time series data using PyTorch.
- ● Accomplished 12% improvement in stitching accuracy while automating multispectral drone imagery conversion into high quality actionable geospatial data for AI-driven Precision Agriculture.
- ● Attained 98% accuracy in overall data quality by implementing a streamlined workflow for stitching, projection, georeferencing, labeling and zonal statistics.
- ● Generated 4.5+ million synthetic datasets across 10,560 scenarios using Python to enhance Agro-ML models.
-
Software Engineer & Cofounder
YVO Service Private Limited (Feb 19 - Jul 21)- ● Achieved seamless user experience with 100 QPS load capacity by spearheading the development of a two-wheel motor service booking system.
- ● Attained 99.5% uptime by engineering and deploying a high-availability NodeJS REST API on Azure, using Docker, SQL and WebSocket.
- ● Enhanced customer engagement and service accessibility by crafting a user-friendly frontend application for a motor-bike service platform, utilizing NodeJs, React, and Tailwind CSS.
- ● Amplified transaction rates by 60% through a data-driven UX redesign, capitalizing on user interaction analytics to optimize and streamline the customer journey.
- ● Managed and mentored a team of five engineers through numerous sprint plannings, facilitating their career advancement and skill development.
-
Graduate Research Assistant
University of Florida (Oct 22- May 23)- ● Performed data preprocessing for plant growth and biomass prediction using Pandas for LSTM model training.
- ● Reduced calibration time by 150 days for DSSAT model, by optimizing Python code implementing parallelism.
- ● Enhanced crop models for 6 farms over 8 years accurately simulating annual soil nitrogen and plant growth data.
-
Teaching Assistant (C++)
University of Florida (Aug 22- Dec 22)- ● Spearheaded a series of lectures on advanced programming concepts such as Lambda Functions, Design Patterns, and Debugging, effectively educating a class of 330 students.
- ● Conducted weekly office hours and doubt sessions to reinforce core programming concepts.
- ● Assisted in developing and grading assignments, quizzes, and exams, with feedback to support student learning.