Sobia Bano —
AI research for a lower-carbon built world.
AI & Data Science researcher at University College Dublin specialising in machine learning, large language models and retrieval-augmented generation applied to energy systems and the built environment. PhD focused on AI-enabled building life-cycle assessment — pairing LLM-powered knowledge systems with predictive modelling and domain-specific datasets.
Bridging artificial intelligence and the built environment — turning fragmented energy & carbon data into decision-ready knowledge.
I'm a PhD candidate at the UCD Energy Institute's Building Energy Informatics group, developing AI-driven methodologies for analysing building energy and carbon performance. My work integrates LLMs and retrieval-augmented generation with structured datasets — EPDs, life-cycle inventories, energy records — to build decision-support systems for sustainable design and retrofit strategy, and applies machine learning to large-scale urban energy performance prediction.
Alongside research, I teach and supervise students in data analytics, programming and applied AI, and mentor MSc students working on AI applications in energy, sustainability and smart buildings. Before my PhD, I spent over a decade as a software engineer and web developer across academia and industry — a background that shows up in the research software I build for the wider building-simulation community today.
Built for where AI is heading next.
Beyond core ML — the agentic, retrieval and knowledge-systems skillset industry teams are hiring for right now.
Where AI meets sustainable buildings.
Four connected threads run through my work — from language models to life-cycle carbon accounting to the software that gets research into practitioners' hands.
AI, LLMs & RAG
Retrieval-augmented generation and LLM-powered knowledge systems that turn scattered technical documents — EPDs, standards, datasets — into queryable, decision-ready tools.
Life-Cycle & Embodied Carbon
Whole-life carbon and embodied-energy assessment for residential retrofit, using structured LCA data to support evidence-based, low-carbon renovation decisions.
Urban-Scale Energy ML
Ensemble machine learning and synthetic building-stock datasets for predicting residential energy performance across entire cities, not just single buildings.
Research Software
Building and maintaining the web platforms — for IBPSA and EC3 — that conference communities rely on to submit, review and disseminate research worldwide.
A decade across academia & industry.
Research roles at UCD sit alongside a long run building real-world software — the combination recruiters in both academia and industry tend to ask about first.
Research Assistant — AI-Driven Building Energy & Life-Cycle Assessment
Academia Jun 2023 – Dec 2025- Conduct AI-enabled building life-cycle assessments (LCA) for energy and carbon analysis
- Develop machine learning and data-driven models for life-cycle GWP evaluation
- Implement LLM and RAG-based approaches for analysing construction and energy datasets
- Data cleaning, validation and preprocessing in Python (Pandas, NumPy, Jupyter)
- Integrate BIM and AI workflows for energy optimisation and sustainable design
Demonstrator
Academia Sep 2023 – Present- Modules: Data Analytics for Engineers, Introduction to Engineering Computing
- Communicate complex technical concepts to non-technical audiences, supporting strong learning outcomes
Web Application Developer — Research Software
Research Software Feb 2021 – Present- Developed and maintained conference paper submission and publication platforms
- Structured data handling (JSON, XML, CSV) and quality control of research outputs
Web Developer
Research Software Mar 2020 – Present- Lead developer for the European Conference on Computing in Construction website
- Supported dissemination of research on digital construction, BIM and sustainability
Private Tutor
Education Sep 2019 – May 2023- Taught Mathematics, Science and Computer Science at Primary and Secondary level (Junior & Leaving Certificate)
Senior Software Engineer
Industry Jan 2017 – Dec 2017- Ecommerce website development and management using WordPress and WooCommerce
Software Engineer
Industry Jan 2013 – Dec 2016- Website development, blogging, Google Play Store management, SEO
- Android app development, live-streaming app development, data analysis
Lecturer
Academia Jan 2011 – Dec 2012- Taught Computer Science at college level; IT Lab in-charge
Doctor of Philosophy — AI-Driven Building Life-Cycle Assessment
University College Dublin Sep 2023 – Present- Develop AI-driven methodologies for analysing building energy and carbon performance
- Integrate LLMs and RAG systems with structured datasets (EPDs, life-cycle, energy data)
- Build decision-support systems for sustainable building design and retrofit strategies
- Apply machine learning models for large-scale energy performance prediction
Bachelor of Engineering — Computer Engineering
NED University of Engineering & Technology Jan 2007 – Dec 2010- Foundations in software engineering, systems design and computer science
Peer-reviewed & conference research.
219+ citations across journal articles and international conference papers on AI-driven building energy and life-cycle performance. Full record on Google Scholar.
Urban building energy performance prediction and retrofit analysis using data-driven machine learning approach
Energy and Buildings, Vol. 303
Urban residential building stock synthetic datasets for building energy performance analysis
Data in Brief, Vol. 53
Intelligent Retrofits in Residential Buildings: A Knowledge-Based Approach
EC3 Conference 2024
A Graph Based Framework to Support Data-driven Urban Building Energy Simulations
EC3 Conference 2024
Residential building energy performance prediction at an urban scale using ensemble machine learning algorithms
European Conference on Computing in Construction, Greece
Data-driven prediction of residential building energy performance at an urban scale through end-use demand segregation
18th IBPSA Conference — Building Simulation 2023, China
Research that ships as usable tools.
From funded national research programmes to the conference platforms thousands of researchers use every year.
ENABLE
Enabling interdisciplinary analysis for whole life-cycle residential renovation. A novel decision-support tool — knowledge base plus web interface — that quantifies scientifically reproducible renovation measures so homeowners, housing agencies and policymakers can make informed, low-carbon renovation decisions.
Urban Building Stock Synthetic Datasets
Synthetic dataset generation covering large-scale residential building stock, enabling machine-learning-based energy performance analysis at city scale as part of the NexSys research programme.
LLM-Powered Knowledge Base for Building Renovation LCA
A RAG-based knowledge system for retrofit decision support, presented at MTU Cork — awarded Best Poster at the IBPSA Ireland Symposium 2024.
Research Conference Platforms
Design, development and maintenance of the submission and publication platforms behind the IBPSA and European Conference on Computing in Construction (EC3) — supporting global dissemination of building-simulation research.
The stack behind the research.
Languages
AI / ML & LLM Stack
Data & Tools
Domain & Platforms
Recognition & credentials.
Fully-Funded PhD Scholarship
Sustainable Energy Authority of Ireland (SEAI)
Best Poster Award — IBPSA Ireland 2024
"LLM-Powered Knowledge Base System for Building Renovation LCA" · MTU Cork
Construction Life Cycle Assessment Specialist
One Click LCA
Building LCA for Architects
One Click LCA
Construction LCA — Embodied Carbon
One Click LCA
Energy Demand in Buildings
Delft University of Technology, via edX
Open to postdoctoral, research & industry AI roles.
Whether it's an academic collaboration, a talk, or an industry role applying AI to sustainability — I'd like to hear from you.