About me
I'm a student at the University of Maryland, passionate about leveraging AI and software engineering to create innovative solutions. My journey in tech is driven by a curiosity to explore the boundless possibilities of machine learning and entrepreneurship. Currently, I'm honing my skills as a Software Engineering Intern at Capital One, focusing on Generative AI applications.
When I'm not coding, you can find me weightlifting, playing basketball, experimenting with new recipes in the kitchen, or spending time with my dog. I'm always looking to learn new technologies and improve my skills.
My projects
NBA Game Predictor
Built a model to predict game outcomes using player and team data, achieving high accuracy.
- Python
- Pandas
- Scikit-Learn
SkyCast: Flight Delay Predictor
Developed an app to predict flight delays using real-time weather data, with a mean absolute error of 1.48 minutes.
- Flask
- React
- Pandas
- Scikit-Learn
University Registration System
Built a scalable course registration system with improved performance and reliability.
- Java
- Multi-threading
My skills
- Python
- Java
- C
- C++
- JavaScript
- MATLAB
- HTML/CSS
- R
- SQL
- Assembly
- React.js
- Node.js
- Pandas
- Scikit-Learn
- Flask
- OpenCV
- TensorFlow
- AWS Services
- Docker
- LLMs
- MySQL
- Jupyter
- Jira
- MongoDB
- UNIX/Linux
- Git
My experience
Software Engineering Intern - Generative AI
Vienna, VA
Developed a Generative AI application tailored for document retrieval and proposal generation by implementing a private Retrieval-Augmented Generation (RAG) system using AWS Bedrock, which improved document retrieval accuracy by 25% and increased proposal acceptance rate by 17%. Implemented efficient processing and management solutions using AWS S3 and Lambda, reducing latency by 11%. Leveraged Jupyter Notebooks for iterative development and testing within an Agile workflow, and utilized Docker for containerization to ensure consistent deployment. Streamlined the CI/CD pipeline design and implementation using AWS KMS and IAM, balancing robust security measures with ease of use.
May 2024 – PresentProduct Innovation Engineer
College Park, MD
1 of 30 students selected for a prestigious 15-month program focused on efforts to improve school safety. Leading the development of an MVP, fine-tuning LLMs with a curated dataset for analyzing textual threats, and integrating TensorFlow for computer vision, resulting in the detection and notification of key personnel about active shooter threats on campuses, backed by a $250K-$2M investment.
Jan. 2024 – PresentFacial Recognition Researcher
College Park, MD
Utilized R to analyze facial recognition data using regression and factor analysis, uncovering significant patterns that improved prediction accuracy by 33%. Identified an 8% accuracy variance in facial recognition across races, genders, and ages, and implemented enhanced algorithmic weighting and adjusted training data sets, reducing bias and improving overall accuracy.
Aug. 2023 – PresentContact me
Please contact me directly at akhil.metukuru2016@gmail.com or through this form.