MVP delivered in 8 weeks.
Funding secured at COP26.
An AI platform that turns thousands of pages of climate policy into structured, searchable insight – built and shipped as a working MVP in 8 weeks, then demonstrated at COP26 to secure funding interest from the World Bank and LSE Grantham Research Institute.
working MVP
for the demo
drive delivery
secured after demo
Client
Climate Policy Radar, a climate technology startup specialising in AI-driven analysis of climate policy and legislation.
Goal
To develop an AI-powered platform that processes and analyses large volumes of climate policy data, enabling policymakers and stakeholders to make faster, data-driven and well-informed decisions for effective climate action.
Climate policy was buried in unstructured text.
Climate Policy Radar’s ambition was to help policymakers make sense of the world’s climate legislation – a sprawling, constantly growing body of documents in dozens of formats. Getting from a stack of PDFs to a decision policymakers could act on meant extracting structured insight from unstructured text, at global scale.
The documents themselves were the obstacle. Policy papers, legislative texts and international agreements varied wildly in structure and format, and none of it was built to be machine-readable. Whatever we built had to handle that variety without losing accuracy, and had to scale from a handful of countries to global coverage – while still giving policymakers information they could trust and act on quickly.
Three constraints were non-negotiable:
- Thousands of unstructured, inconsistently formatted policy documents
- No existing way to extract structured insight from raw text
- Had to scale from national to global policy data
A semantic search engine built for climate legislation.
We built and structured Climate Policy Radar’s data science and engineering teams, then designed the overall technical strategy for the AI platform. That strategy started with the hardest part of the problem: getting clean, structured text out of policy documents that were never designed for machines to read.
Policymakers didn’t need more documents. They needed a platform that had already read them.
AI-driven PDF text extraction and document layout analysis turned raw policy documents into structured data. On top of that, we built a semantic search and retrieval system using Sentence Transformers and Elasticsearch, so policymakers could search by meaning rather than exact keyword matches.
The result was an MVP of the Climate Policy semantic search tool, delivered in 8 weeks. It was demonstrated live at COP26, in front of the organisations who would go on to back it.
From MVP to COP26 funding in 8 weeks.
Climate Policy Radar went from an idea to a working AI platform, and from a working platform to funded and credible, inside two months. The semantic search tool gives policymakers faster, more accurate access to the climate legislation they need to act on, and the COP26 showcase turned that capability into strategic partnerships that support the platform’s long-term growth.
* Case studies reflect work undertaken by our Heads of AI either during their tenure with Head of AI or in prior roles before they were part of the Head of AI network; they are provided for illustrative purposes only and are based on conversations with our Heads of AI.
More case studies.
Your biggest pain point.
Fixed in 14 days. 50% off.
This started with one conversation. Book a 30 minute brainstorm call — we’ll plan your first AI project together and issue your 50% discount code. No payment today.
*Case studies reflect work undertaken by our Heads of AI either during their tenure with Head of AI or in prior roles before they were part of the Head of AI network; they are provided for illustrative purposes only and are based on conversations with our Heads of AI.