Danyal Aftab
I am a Software Engineer and AI researcher currently based in Ireland, with experience spanning industry, academia, and applied research.
About
I am a Software Engineer and AI researcher currently based in Ireland, with experience spanning industry, academia, and applied research. I am presently working at Technological University Dublin, where I focus on large language models, model compression, and system-level AI integration. My research includes Tailored-LLaMA, a structured pruning and task-specific prompting framework that retains high accuracy while significantly reducing model size, which I presented at ECAI 2024.
Previously, I worked as a Software Engineer at Teradata, contributing to ModelOps platforms using Spring Boot, Angular, Docker, and PostgreSQL, and earlier as a Machine Learning Engineer at Vidizmo, where I optimized deep learning models for video analytics. I have also completed a software engineering internship at Autodesk, building AI-driven sustainability tools that integrate large-scale environmental datasets to support low-carbon construction decisions.
My interests lie at the intersection of LLM efficiency, applied machine learning, and real-world AI systems, with multiple peer-reviewed publications in top venues including IEEE TIP, CVPR Workshops, and ECAI. I enjoy translating advanced research into practical, scalable solutions with measurable impact.
- 3 years' experience
- Joined December 2024
Industry experience
- Software & Internet
- Media & Entertainment
- Education
- Professional Services
Work experience· 7 roles
Designed and deployed a high-performance Model Context Protocol (MCP) server for Anyscore enabling secure, real-time access to sports analytics data; integrated MCP server with LibraChat MCP client configured with the Azure GPT-5 API to support natural-language data queries; automated data-analysis workflows, reducing processing time by over 75% and enhancing analytics efficiency.
Designed and deployed a high-performance Model Context Protocol (MCP) server for Anyscore, enabling secure and real-time access to sports analytics data. Integrated the MCP server with LibraChat MCP client using the Azure GPT-5 API to support natural-language queries over complex datasets. Enabled advanced match-insight generation including player performance metrics, team statistics, and tactical breakdowns, increasing analytical depth. Automated data-analysis workflows, reducing processing time by more than 75% and improving efficiency across the analytics team.
Implemented an MCP server to enable LLMs to access real-time environmental data for construction materials; built an Idemat connector within the Sustainability Data API handling 3,000+ Environmental Product Declarations; reduced material sustainability research time from hours to seconds, enabling data-driven carbon footprint analysis and low-emission material selection.
Education
Certifications and awards· 9
Hire Danyal Aftab today
Post a job and invite Danyal, or contact them directly. No hidden commission.