I’m an AI developer specializing in secure, scalable systems and LLM applications. My work bridges research and production, with a focus on real-world impact and user experience.
/ My StoryHow I started as an AI developer
My journey into AI began with a curiosity for solving real-world problems using data and code. I started with small projects like object detection and quickly moved into building full-stack AI applications, ranging from chatbots to investment advisors. Over time, I realized the power of AI to make decisions smarter, faster, and more secure. Now, I build systems that combine LLMs, automation, and analytics to drive real impact.
/ My StoryMy first AI related work
In 2023, I developed an object detection system that identified and measured bruises on fruits using Detectron2. The goal was to automate quality control and reduce produce waste through more accurate damage assessment. I trained the model on custom-labeled data and fine-tuned it for precision across different fruit types. That project laid the foundation for my interest in AI and its potential to solve practical, real-world problems.
LanguagesPythonJavaSQLCJavaScriptTypeScript
Developer ToolsGitVS CodePyCharmJupyter
LibrariesTensorflowPyTorchCUDATransformersHuggingFaceLlama.cppNumPyReactNext.jsPandasPyMongo
PracticalLarge Language Model (LLM)Natural Language Processing (NLP)Retrieval Augmented Generation (RAG)Vector DatabaseNo-SQL databaseComputer VisionMachine LearningArtificial Intelligence
MostEdgeEngineered a Retrieval-Augmented Generation (RAG) pipeline powered by a private LLM to analyze market trends by integrating data from multiple SQL databases. Optimized context handling and memory management to ensure accurate, multi-turn responses. Designed and deployed a responsive user interface that enables users to query the LLM using natural language and apply contextual filters for refined analysis.
MAY 2025 / AUG 2025

Georgia Tech
Conducting research on hallucinations in various Large Language Models (LLMs) and their impact on user trust and decision-making. The work involves categorizing different types of hallucinations, analyzing the reliability of model responses, and identifying patterns in model behavior. Additionally, exploring and testing counter-measures aimed at reducing hallucinations to enhance the accuracy, safety, and overall user experience in AI-assisted interactions.
JAN 2025 / AUG 2025
Georgia Institute of Technology / MAY 2024 – MAY 2027
Dean’s List student focused on AI systems, large language models, and data-driven security solutions.