The team at POP approached us to help them with soem significant data parsing and amangement challenges with their Belong in Plymouth project. They had two separate instances where they had huge amounts of data that they needed to be analysed, organised and then made accessible via their website.
Conversations
The BIP team spoke to hundreds of Plymouth residents — in cafés, community halls, quiet corners — and gathered an extraordinary collection of real conversations. We used AI to index and archive these conversations, breaking them into fragments and themes.
Instead of a static archive, we created a searchable dialogue. Anyone can now visit the website and ask questions in plain language — like “What do people say about feeling safe?” — and explore real responses from real people.
Journal
Alongside this, we captured the inner workings of the Belong in Plymouth project itself. Over 3 years of emails, meeting notes, reflections and updates were distilled into over 100 short blog posts using a combination of AI and Pinecone, a vector database solution.
Website & Airtable CMS
We built a custom website using Astro, our favourite Javascript framework. This was paired with a simple intuitive content management system powered by Airtable which allowed the POP team to easily upload conversations, have AI parse and tag them, and then moderate the results.