Citypress needed an AI platform that would give them access to multiple LLM models while maintaining complete control over client data. Ortom helped design and deploy a custom solution on their own infrastructure, enabling AI adoption across previously restricted client accounts.

Challenge

Citypress, a Manchester based PR and communications agency, faced a growing challenge. AI adoption was accelerating rapidly, both within their industry and internally among their team. While the agency had invested in an enterprise LLM license, they encountered critical limitations that prevented them from using AI tools effectively for many clients.

Rob Stone, who heads up Innovation and Digital at Citypress, explains: "AI use was increasing rapidly, both in our industry with our competitors, but obviously internally within the agency as well. And we knew there were a few limitations to us being able to use some of these tools for our clients."

The standard enterprise AI platforms lacked crucial features. Citypress couldn't save and control their own data in ways that met client contractual obligations. Some clients didn't allow data to be processed or stored in certain locations—requirements that off-the-shelf enterprise licenses couldn't accommodate. Additionally, the agency wanted flexibility to access different AI models rather than being locked into a single provider in a rapidly evolving landscape.

Citypress needed a bespoke solution: a platform they owned and controlled, hosted on their own infrastructure, with access to multiple LLMs while meeting strict data governance requirements. Finding the right partner proved challenging. "We spoke to a lot of people where the approach wasn't very flexible," Rob recalls. "They tried to shoehorn us into whatever platform they were using." Many developers were pushing their own pre-built platforms rather than listening to Citypress's specific requirements.

After exploring options including their freelance network and conducting online searches for local AI developers, Citypress connected with Ortom.

Solution

From the initial conversations, Ortom demonstrated a different approach. Rather than proposing an existing platform, Ortom took time to understand Citypress's specific constraints and business context. The brief was clear: develop a platform that Citypress owned, giving them access to multiple LLMs alongside complete control over their data and its location.

The solution was based around on AWS infrastructure that Citypress controlled entirely. This was integrated with their Microsoft Entra system for seamless and secure user management. LibreChat was used as the front-end interface, using with AWS Bedrock to provide access to multiple AI models. Critically, all data—including documents, conversations, and client information—remained within Citypress's own AWS environment and stayed within the EU to meet data governance requirements.

This architecture meant Citypress could now use AI tools on client accounts where previously it had been contractually prohibited. The platform offered flexibility to switch between different LLM providers as the technology evolved, protecting their investment against rapid market changes.

The process began with Tom visiting Citypress's office to meet senior stakeholders. "I think at that point we probably still needed to convince them that this is something we wanted to do," Rob notes. "After meeting you, they were very enthusiastic about taking this project on and getting it done." Throughout development, Ortom maintained close communication with multiple stakeholders across the agency, adapting to their needs while suggesting additional features that would enhance the platform's capabilities.

"It was very obvious to us talking to you from the outset that you really knew what you were talking about, that you had a very good understanding of the challenge we were trying to overcome... For us, that's kind of how we've seen it. It's definitely been a partnership over the last few months, and I think you get much better results that way."

Rob Stone
Head of Innovation and Digital, Citypress

Results

• Enabled AI Use on Previously Restricted Accounts: For the first time, Citypress teams could use AI tools on client accounts where contractual data requirements had previously made this impossible, improving efficiency and scale of work.

• Delivered Multi-Model Flexibility: The platform provided access to multiple LLM providers, allowing team members to experiment with different models and develop preferences for specific tools suited to different tasks.

• Achieved High Internal Adoption: Uptake of the platform across the agency was notably strong, with teams quickly integrating AI capabilities into their workflows.

• Maintained Full Data Control: All client data remained within Citypress's infrastructure and EU-based servers, meeting strict contractual and governance obligations while enabling AI adoption.

• Protected Against Technology Lock-In: The flexible architecture ensured Citypress wasn't dependent on any single AI provider in a rapidly evolving market, protecting their investment.

• Delivered On Time and On Budget: The project completed within agreed timescales and budget, with no significant issues during implementation.

"The uptake on the tool has been really high. On clients where we weren't able to use AI at all, really, especially when it utilized any of their data, we're able to do so for the first time. The people that work on those accounts are definitely really enjoying the ability to use these tools to make their work more efficient, to do things at greater scale."

Rob Stone
Head of Innovation and Digital, Citypress

Conclusion

Citypress now operates a secure, flexible AI platform that meets the demanding data governance requirements of their client base while providing teams with access to cutting-edge AI capabilities. The solution has removed a significant barrier to AI adoption across previously restricted accounts, enabling the agency to work more efficiently at greater scale.

The platform positions Citypress well for future AI developments. With plans already underway for Phase 2 enhancements to support specific client use cases—including more sophisticated implementations for major clients like Lloyds Banking Group and custom chatbot deployments—the foundation we built together provides a robust basis for continued innovation.

This project demonstrates Ortom's approach to working with non-tech-first businesses. By taking time to understand the specific constraints of the communications industry and working collaboratively with multiple stakeholders, we delivered a practical solution that balanced technical sophistication with real-world business requirements. The result is a platform that solves today's challenges while remaining flexible enough to adapt as both AI technology and client needs evolve.

"The delivery that we've got, the work that we've had done, and the impact that it's made internally has been really significant, really positive, and it's been a real joy to work with you, Tom... We'd definitely be happy to recommend you to others—as long as it didn't take up so much of your time that we weren't able to get hold of any of it ourselves."

Rob Stone
Head of Innovation and Digital, Citypress