Keyah Consulting and Sustained Futures are running the Real Estate Build Studio, a live course that teaches sustainability leads to build an AI agent for real estate sustainability disclosures. The agent reads asset-level reports, including EPCs, and turns them into one comparison table with a page reference for every value. The course is led by Dr Sophie Taysom, Founder and CEO of Keyah, and Katie Neck, Founder of Sustained Futures. It runs on Zoom on 18 and 25 November 2026, 4.30 to 6pm UK time.
Why are real estate sustainability disclosures hard to compare?
Real estate disclosures are inconsistent, patchy and hard to compare. Two assets can both report their emissions and still produce figures that cannot be set side by side, because they use different emissions boundaries, floor area measures or reporting years.
Pulling energy, emissions, certifications, targets and climate risk data out of each report by hand is slow. It is also easy to miss where figures only look comparable.
How can AI support sustainability leads?
An AI agent is a set of instructions written once, so the AI follows the same steps every time it is given new documents. For a disclosure review, it reads each report, extracts the same data points from every asset and places them in a single table.
The agent built in this course has checks designed in:
- every value carries a page reference, so it can be verified
- anything missing is marked “not disclosed” instead of guessed
- figures that cannot be compared are flagged, such as different emissions boundaries, area measures or reporting years
The first pass takes minutes. That frees sustainability leads to spend their time on the part of the work that needs expertise.
Why does the translation layer still matter?
A table is not a decision. Someone still has to decide what it means, check that it is right and judge what is safe to conclude before it informs an investment, lending or board decision.
That translation layer is the focus of Keyah’s work. Dr Sophie Taysom helps real asset owners, investors and their advisors turn frameworks, regulation and data that do not line up into something they can act on. The course covers where human judgement has to step in, and what to do before the table informs a decision.
What will participants build?
Participants build and run a live agent on a sample portfolio of four assets: an office, a logistics warehouse, a mixed-use building and a build-to-rent scheme. The data points follow the GRESB Real Estate Assessment indicators.
- Session 1, 18 November: build the agent live, then try it on public reports between sessions
- Session 2, 25 November: review the output, troubleshoot common issues, add a new data point and cover how to use the agent in practice
Everyone leaves with a working agent, the exact prompt, a step-by-step tutorial and a repeatable method for building agents for other purposes.
Who is the course for?
- Sustainability leads comparing disclosures across a portfolio, fund or supply chain
- Asset managers checking what each asset reports on energy, emissions and climate risk
- Surveyors reading EPCs and asset-level reports at volume
- Advisory and engineering professionals working across real asset portfolios
No prior AI or coding knowledge is needed. Participants need a paid Claude Pro account.
Course details
- Dates: Wednesday 18 and 25 November 2026, 4.30 to 6pm UK time
- Format: two live 90-minute sessions on Zoom, both recorded
- Cost: £360 + VAT for both sessions
- Book: Real Estate Build Studio
An employer sign-off letter is available on the booking page.
About the trainers
Dr Sophie Taysom founded Keyah Consulting, a London-based climate and sustainability advisory firm, in 2017. She contributes to the Financial Times’ Sustainable Views and PERE, has presented to the Association of British Insurers and a UK Parliamentary roundtable, and founded the 10,000+ member Sustainable Real Estate LinkedIn group.
Katie Neck founded Sustained Futures, a Future Skills social enterprise and 2026 Earthshot Prize nominee. She has led AI upskilling for more than 100 impact leaders through the Sustained Futures AI Accelerator.
Frequently asked questions
What is an AI agent for sustainability disclosures?
It is a set of instructions that tells an AI tool how to read sustainability reports, which data points to extract and how to check its own output, so it follows the same method on every document.
Do I need to know how to code?
No. The course assumes no prior AI or coding knowledge.
What software do I need?
A paid Claude Pro account, set up before the first session.
What if I cannot attend live?
Both sessions are recorded.
Can the method be used for other tasks?
Yes. Participants learn a repeatable method for structuring instructions, building in checks and testing output, which they can use to build other agents.
Does the AI replace expert judgement?
No. The agent produces a verifiable first pass. Deciding what it means for a portfolio, and what is safe to conclude, stays with the expert.