Crowdfunder is a UK-based platform that helps individuals, charities, and businesses raise money for their projects. Their platform connects project creators with both retail and enterprise donors. As the company has grown, so has the importance of data in helping it scale. They wanted better ways to manage, understand, and use their data, whether to better match donors to the right projects or power internal reporting.
To do that, Crowdfunder decided to explore moving from Amazon Web Services (AWS) to Google Cloud Platform (GCP). They were especially interested in using services like BigQuery and Google’s generative AI tools to make their platform smarter and more efficient. So our team joined them on this journey, starting with a proof of concept focused on BigQuery and generative AI using Gemini.
The challenge
Crowdfunder’s existing platform was based on several AWS components that helped them manage web traffic, store data, and deploy updates. Their web app ran across multiple servers that could scale with demand, and they used modern tools to handle background tasks, databases, and caching. A separate system powered their content management, and they also used services to secure traffic and monitor performance.
While their existing tools were stable and scalable, Crowdfunder saw an opportunity to improve how their teams accessed and used data. They were particularly interested in Google Cloud’s data and AI tools to improve visibility across the organization and explore how AI could help with specific problems, like refining the application approval process, where traditional categorization methods fell short.
Their biggest challenge was data. They collected a lot of it — like project information, donor interactions, and user activity — but their access was limited. The main data warehouse was managed by an external partner using Amazon Redshift, which meant their internal teams had limited visibility and couldn’t easily run their own reports or experiments. Nightly ETL jobs pushed data from a read-only MariaDB instance and Segment into Redshift.
The company wanted more reliable access to their data, ideally through a more modern platform like BigQuery. They also wanted to explore practical uses for generative AI to reduce manual work, such as how AI could help categorize projects, match large donor funds to the right campaigns, or even assist with writing descriptions.
“We didn’t have the level of access to our data that we wanted. We knew it was there, but it wasn’t easy for the team to explore or use it day to day. Being able to change that, and even test out things like matching projects to funding with AI, was really exciting for us.”
— Nikolai Clark, Head of Platform, Crowdfunder
As the company grew, Crowdfunder saw the opportunity not just to switch platforms, but to modernize their whole infrastructure. The cost and complexity of their existing AWS setup pushed the team to explore alternatives — ideally something more cost-efficient and easier to manage.
The solution
We worked with Crowdfunder to explore how BigQuery and generative AI could deliver value in real-world use cases.
Data migration to BigQuery
The first step was moving key tables that hold data on crowdfunding projects, transactions, and fund applications into BigQuery. Together with their team, we set up a process to regularly export this data and load it into BigQuery. After an initial full upload, the system was set up to automatically update the data every day.
With the data in BigQuery, Crowdfunder could start exploring it directly, running useful queries and seeing how well BigQuery could work as their main data platform in the future.
BigQuery as a replacement for Redshift
A major part of this project was helping Crowdfunder consider BigQuery as a replacement for Redshift. We ran sessions on optimizing queries, understanding cost structures, and managing access with tools like row- and column-level security. We also showcased how to manage transformation logic using SQL inside BigQuery and more advanced ELT orchestration with Dataform.
“Moving away from AWS was a big decision, especially with core systems like our database and data warehouse. Revolgy helped us see how Google Cloud, mainly BigQuery, could handle our data needs. The ability for BigQuery to deal with large data at speed is fantastic; we now churn all our historic data to produce our BI tables in just one minute.”
— Phil Geraghty, Co-founder & Director, Crowdfunder
Matching projects to donor funds with genAI
Another real challenge for Crowdfunder was automatically matching the right projects with available donor funds. This involved understanding project descriptions, categories, and other key details to determine which campaigns best fit each funding opportunity.
We explored two use cases — automatically categorizing projects and, more importantly, matching projects to specific funding criteria. For the matching task, we used Gemini directly within BigQuery. We built queries that could read and summarize project text, score how well each project fit with different funds, and even explain why a match was a good one — all within the same workspace Crowdfunder used for their data.
To test the setup, we ran this process on more than 9,000 projects. While the initial bulk run took some time due to usage limits, we designed the system for fast, daily updates going forward, making it a practical and affordable solution.
Results & improvements
- Crowdfunder has access to their core data in BigQuery, so the team can run their own reports and explore ideas without relying on an external partner.
- The AI proof of concept helped automate project-to-fund matching, saving time and showing how AI can support day-to-day work.
- The system ran successfully even with over 9,000 live projects. It now updates daily, efficiently, and at low cost.
- We walked the team through how the setup works and outlined the next steps for moving it into production instead of just handing over a technical solution.
- Crowdfunder now has a clearer understanding of their data, a working AI example, and a roadmap they can build on.
“The Crowdfunder team knew they wanted better data access, practical AI applications, and efficient infrastructure. We designed the cloud future they imagined and tested the most demanding use cases directly on their real data. The project team was engaged and collaborative throughout, and we’re happy that our AI proof of concept delivered very conclusive results.”
— Lubomír Dočkal, Data & AI Domain Architect, Revolgy
About Crowdfunder
Crowdfunder is a fast-growing UK platform that enables fundraising for social causes, personal projects, and community development. Their customers include individuals, NGOs, and businesses. They bring together a network of donors, including strong support from enterprise partners, and help people turn funding ideas into action.
Crowdfunder has a team of around 45 people and is growing at about 20% annually. With an increasing need to scale efficiently, they see technology — and AI’s role in particular — as critical to their future.
Conclusion
Revolgy helped Crowdfunder learn more about how Google Cloud could support their long-term goals, starting with a data-focused proof of concept using BigQuery and Gemini. The project gave the company better access to their data, more control, and a working example of how AI can make everyday tasks faster and easier. They continue exploring their next steps with Google Cloud, while Revolgy supports them on the data and AI side.
At Revolgy, we help teams move to Google Cloud, modernize their data stack, and explore real-world AI applications. If you’re looking to do the same, get in touch — we’d love to help.
