Associate AI/ML Engineer
đ§ dhilshanroney@gmail.com | đ +94 76 332 2617 | đ Colombo, Sri Lanka
âšī¸ LinkedIn | đ Portfolio | đž GitHub
PROFILE
Driven Data Science graduate with hands-on experience as an AI/ML Engineer Intern. Passionate about applying AI and Machine Learning principles to real-world challenges, with a focus on delivering impactful, data-driven insights on time and within scope. Eager to contribute to innovative AI projects, collaborate with teams, and grow professionally in a forward-thinking organization dedicated to advancing Artificial Intelligence and Data Science.
TECHNICAL SKILLS
-
Machine Learning:
Regression, Classification, Clustering, Neural Networks, Deep Learning, Reinforcement Learning, Autoencoders, Hyperparameter Tuning, Evaluation Metrics
-
AI Techniques:
LLM (Large Language Models), OpenAI API, LangChain
-
Data Analysis:
Data Cleaning, Feature Engineering, EDA, Model Training
-
Languages & Libraries:
Python, R, SQL, HTML, CSS, Bash/Shell, TensorFlow, Keras, Scikit-learn, Pandas, NumPy, Matplotlib, Seaborn, Hadoop, PySpark, HuggingFace, FAISS
-
OS & Tools:
Linux, Jupyter Notebook, VS Code, Google Colab, Git/GitHub, Power BI, Tableau, Excel, Ollama, Selenium
-
Databases:
MySQL, MongoDB
-
Web Development & Visualization:
Flask, Streamlit, FastAPI
WORK EXPERIENCE
AI/ML Engineer Intern | November 2025 â May 2026
FAITE â IT Services and IT Consulting, Jaffna
- Worked as part of the team to design and build AI/ML Architectures for end-to-end projects, ensuring scalability, efficiency, and alignment with business objectives.
- Participated in Client Discussions to gather requirements and understand business needs, ensuring the development of tailored AI solutions that align with client expectations.
- Collaborated with a team of AI Developers, Software Engineers, and UI/UX Design Engineers across all projects to design, build, and deploy AI/ML Architectures.
- Developed and implemented Machine Learning Models to solve real-world problems, resulting in a 15% improvement in prediction accuracy.
- Worked on Neural Network Models for data classification, image recognition, and natural language processing, improving processing speed by 20%.
- Conducted Hyperparameter Tuning on multiple models, reducing training time by 25% while maintaining model performance.
- Applied FastAPI to build and deploy scalable machine learning APIs, reducing deployment time by 30%.
- Utilized Large Language Models for NLP tasks, enhancing text classification accuracy by 18%.
- Worked on developing Convolutional Neural Networks for image recognition tasks, improving classification accuracy by 22%.
- Designed and implemented Chatbot Solutions using NLP, improving user engagement by 30%.
- Led Data Cleaning and Feature Engineering processes, improving data quality and reducing model errors by 22%.
- Employed Evaluation Metrics to assess model performance, ensuring alignment with business objectives and KPIs.
- Maintained Version Control in GitHub, ensuring smooth collaboration across a distributed team of 10+ developers.
KEY PROJECTS
01. CCTV Camera Detection for Supermarket (Canada)
- Developed an AI system for detecting the status of cashier drawers (opened/closed) and integrated it with the Point of Sale system.
- Detecting unrelated behavior between Customers and Cashiers to improve supermarket security.
Tech stack: Python, OpenCV, TensorFlow, Keras, FastAPI, POS Integration, Ultralytics YOLO
02. AI Marketing (Canada)
- Worked on Real-Time Customer Behavior Tracking, analyzing shopping habits to optimize in-store experiences.
- Developed Hyper-Personalized Customer Experiences and Dynamic Pricing Models to increase sales and customer satisfaction.
- Built AI-Powered Chatbots and Virtual Assistants to handle customer queries and improve service efficiency.
- Developed Customer Lifetime Value Prediction models to identify valuable customers.
- Designed an AI YouTube Promotion Tool that optimizes video content delivery based on viewer behavior.
Tech stack: Python, TensorFlow, Keras, FastAPI, GCP, NLP, Chatbot Development
03. Cybersecurity (UK)
- Led the development of an AI system for False-Positive Reduction in threat detection systems.
- Worked on Clone Website Detection, Suspicious Script Analysis, and Behavior-Based Malware Detection.
- Designed a system for Overall Risk Scoring (0â100) and conducted Industry-Based Risk Comparison.
- Built an AI-Based Threat Prediction system that identifies anomalies and predicts potential breaches.
- Conducted Credential Leak Analysis to ensure data privacy.
Tech stack: Python, Scikit-Learn, TensorFlow, Keras, OpenCV, GCP, NLP, FastAPI, Cloud Security Tools
04. OCR Recognition for UK School (Secondary Level)
- Developed an OCR Grading System to compare answer sheets with marking scheme PDFs for subjects like Maths, English, and Science.
- Automated grading of graphs, multiple-choice questions, written answers, and mathematical equations, reducing time spent on manual evaluations by 40%.
Tech stack: Python, OpenCV, Tesseract OCR, TensorFlow, PyTorch, HuggingFace
PERSONAL PROJECTS
Resume Categorization System
- Developed a scalable Resume Categorization System to categorize resumes into specific job roles, preprocess text data, and create an interactive web app for users to upload and organize categorized resumes.
Tech stack: Streamlit, Pandas, Pickle, ML Models, docx2pdf, pypdf
AI Web Scraper â Scrape ANYTHING
- Developed an AI-powered web scraper using Streamlit, Selenium, BeautifulSoup, and LangChain, enabling users to scrape and parse website content with dynamic AI-driven extraction based on custom queries.
Tech stack: Python, Selenium, BeautifulSoup, LangChain, Ollama, ML Models
Rainfall Prediction Using Python & MLFlow
- Developed a Rainfall Prediction application using Flask, MLflow, and machine learning models to predict rainfall based on meteorological data, providing real-time predictions via a user-friendly web interface.
Tech stack: Flask, MLflow, Pandas, ML Models
Agentic Climate Early Warning System for Sri Lanka
- Developed a multi-horizon precipitation forecasting model using Spatio-Temporal Graph Convolutional Networks to predict rainfall 1â7 days ahead.
- Aggregated station-level forecasts to district-level risk classifiers for flood/drought predictions.
- Implemented autonomous alerting with SHAP-based explanations for district-level decision-making.
Tech stack: Python, Streamlit, ML Models
Chat with Multiple PDFs â LangChain App
- Developed a MultiPDF Chat App that uses a language model to answer questions based on the content of multiple PDFs, leveraging text extraction, chunking, and similarity matching.
Tech stack: LangChain, FAISS, pypdf, OpenAI API, HuggingFace API, ML Models
For more projects and information: github.com/Dhilshan2000
EDUCATION
-
BSc (Hons) in Data Science | 2022 â 2025
National Innovation Centre (NIBM), Colombo, Sri Lanka (Affiliated with Coventry University, UK) Overall Classification: Second Class, Upper Division
CERTIFICATIONS
- Generative AI Fundamentals, Databricks
- Data Science Orientation, IBM
- Introduction to Generative AI, Google
- CNNs with TensorFlow: Basics of Machine Learning, Coursera
- Exploratory Data Analysis for Machine Learning, IBM
LANGUAGES
- English (Fluent)
- Tamil (Native)
- Sinhala (Conversational)
REFERENCES
- Sanjeth Srikathirkamanathan
- Position: Assistant Director, AI Data & Digital â Acuity Analytics
- Tel: +94 11 235 6295, Mobile: +94 75 571 8427
- Email: sanjeth.srikathirkamanathan@acuityanalytics.com
- Address: Level 32, West Tower, World Trade Center, Echelon Square 1, Colombo 01, Sri Lanka
- Dhushaker Yogeswaran
- Position: Senior Systems Engineer â IFS R&D International
- Mobile: +94 76 838 1848
- Email: dhushiymost@gmail.com
- Address: Level 18, Orion Towers, 736 Dr Danister De Silva Mawatha, Colombo 09, Sri Lanka