Dilaksan Thirugnanaselvam

AI Research Engineer

Contact

Email thirudilak131@gmail.com Phone +94 0788702104 Location Batticaloa, Sri Lanka

Languages

100% English

100% Tamil

Expertise and Competencies

Machine Learning 0%

Deep Learning 0%

NLP 0%

Generative AI 0%

Tools & Technologies

Python Java C++ TensorFlow PyTorch Scikit-learn PennyLane Qiskit Cirq GPT-4 BERT RAG Fine-tuning Prompt Tuning LangChain LangGraph LangSmith n8n ADK Streamlit SQL AWS FastAPI PostgreSQL Docker GitHub VS Code Google Colab Jupyter Notebook

Hello, I'm Dilaksan!

As a motivated AI Research Engineer with strong skills in Python and mathematics, I am passionate about Generative AI, AGI, and Quantum AI. I strive to build transformative technologies that enhance decision-making, boost efficiency, and revolutionize experiences, while continuously learning and exploring cutting-edge innovations. Engaged in AI research to develop impactful technologies.

Contact Me

Projects

Research Projects

LinkedIn Followers

Certifications

Education

University of Vavuniya

August 2021 – August 2025

BSc (Hons) in Computer Science

Relevant Coursework: AI, Machine Learning, Data Structures, Algorithms, Web Development, Operating Systems, Software Engineering, Probability, Statistics, Linear Algebra, Research Project

Experience

Artificial Intelligence Engineer Intern

Oct 2024 – Dec 2024

StudentPod

Developed AI models using NLP for personalized career recommendations, integrated into a real-time Streamlit interface.
Enhanced model accuracy by optimizing algorithms and using feedback loops to continuously improve performance.

Associate AI Engineer

Jan 2025 – Sept 2025

Detech

Assisted in data preprocessing, model training, fine-tuning Large Language Models (LLMs), and optimizing performance for AI/ML projects.
Contributed to AI research, development, and deployment of AI-driven applications while collaborating with the team on documentation, debugging, and process improvements.

Artificial Intelligence Research Assistant

Feb 2025 – Jul 2025

DreamSpace Academy

Specialized in LLMs, Generative AI, and Machine Learning, focusing on model fine-tuning and real-world applications; actively working on research paper publication.
Built and deployed ML and Deep Learning models, optimizing performance with advanced techniques and seamless integration into real-world applications.

Research Assistant

Oct 2025 – Present

Sabaragamuwa University of Sri Lanka

Conducted research and development in generative AI, deep learning, and quantum machine learning.
Collaborated with undergraduates on AI research and model development projects.

Projects

Chatbot Application

Chatbot powered by LLaMA-2, developed in Python using Streamlit for intelligent interactions and a user-friendly interface.

LLaMA-2 Python Streamlit

Mobile Price Range Prediction

Interactive dashboard using Streamlit, integrating TensorFlow-based ANN for dynamic mobile price predictions.

TensorFlow ANN Streamlit

Airline Sentiment Analysis

Preprocessed data, extracted features using TF-IDF, and trained Naïve Bayes and Random Forest models with a Streamlit app for real-time sentiment analysis.

TF-IDF Naïve Bayes Random Forest Streamlit

Recipe Rendezvous

App powered by a fine-tuned GPT-4 model, developed using Python and Streamlit, generating creative recipes based on available ingredients for chefs and home cooks.

GPT-4 Python Streamlit

PathfinderAI

Intelligent career recommendations using user data, developed in Python with a Streamlit interface and leveraging the OpenAI API for analysis.

OpenAI API Python Streamlit

Scholarship News Analytics & Deadline Tracker

Uses GPT API to analyze scholarships and push updates to a Telegram channel, helping users track deadlines and discover opportunities with real-time notifications.

GPT API Telegram Python

Love AI Agent

Built using LLaMA 3.3, it generates poetic love messages, romantic advice, and emotional support with vibrant animations and glowing text for an enchanting user experience.

LLaMA 3.3 Python NLP

GraphRAG

Implemented from scratch with support for incremental indexing, vector-based querying via pgvector, community detection using the Leiden algorithm, PostgreSQL integration, and containerized deployment with Docker.

Python PostgreSQL Docker

Quantum AI Agent: Binary Iris Classifier

Developed a Hybrid Quantum-Classical Neural Network (QNN) with a Gradio interface for real-time binary classification and batch predictions, gaining hands-on experience in Quantum AI and PyTorch.

Quantum AI PyTorch Gradio

Murugan Bot

Built a culturally grounded Murugan-themed AI chatbot using RAG, LangChain, and Streamlit to deliver context-aware responses rooted in Tamil traditions.

RAG LangChain Streamlit

Research Projects

Investigate LLMs and Develop a Model for Numerical Linear Algebra

Nov 2024 – Present

Evaluated LLMs like o1-preview and GPT-4o on linear algebra, fine-tuned models for abstract reasoning, leveraged FrontierMath benchmarks, and advanced AI reasoning for solving complex, real-world problems in high-precision mathematical domains.

Certifications

Become an AI Engineer: LinkedIn Learning Path
IBM AI Engineering Professional Certificate: IBM
Become an OCI Generative AI Professional: Oracle University
AI/ML Engineer - Stage 1: SLIIT
Azure Fundamentals: Microsoft Learn

Cloud Concepts: Microsoft Learn Student Ambassador Event
Google Analytics: Google (Beginners & Advanced)
Python for Data Science: IBM
Getting Started with Deep Learning: IBM
LLMs: Udemy