Twimbit is a tech-forward research and advisory firm that harnesses the power of AI. It delivers a private, contextual platform to its clients, helping them excel as leaders in their domain by providing access to industry insights in various formats.

At Twimbit, I was the sole UI/UX designer throughout my tenure. My role involved working closely with the founders to understand their vision for the company’s rapidly evolving landscape. They wanted to build insight-driven products that kept the users at the heart of it all.

I led the design and experience strategy for Buddy—an internal AI productivity tool that evolved into a potential standalone product. My key responsibilities included:

Product Concept Development : Took ownership at the conceptual stage and helped the project grow beyond its original scope.

User Research : Conducted internal user interviews to identify pain points in the CMS workflow and understand hesitations around AI adoption.

UX Strategy & Feature Definition : Translated research insights into feature requirements so Buddy could effectively serve the analysts’ real needs.

End-to-End Product Design : Designed the platform interface to integrate multiple AI models and embed Twimbit’s business insights seamlessly.

Cross-Functional Collaboration : Worked closely with the AI and development teams, as well as stakeholders, to ensure alignment and maintain clear communication throughout the process.

While I was working at Twimbit, there was a requirement to build something unique and useful for the internal team—leveraging AI to improve efficiency. The project was originally intended to be part of the internal CMS infrastructure, designed to encourage faster content creation and reduce friction between writing and publishing. The leadership team brainstormed what this tool would include and called upon the wider team to support its development.

I led the design for the tool that retained its working title, Buddy—a powerful AI chatbot platform that aggregates leading AI models such as ChatGPT, Claude, Perplexity, and others into one seamless interface. The platform combined multi-model capabilities with Twimbit’s proprietary industry insights, creating a one-stop solution for advanced AI interaction and business intelligence. It also featured built-in and customizable assistants that supported common workflows, while allowing users to create tailored ones to meet their specific needs

Once we developed a prototype, stakeholders quickly began to see its potential as a client-facing product. What started as an internal tool was soon being considered as a standalone offering—an all-in-one solution for the distributed AI landscape, providing a unified interface to access leading models while being enhanced by Twimbit’s research and insights.

What began as a simple concept quickly became my responsibility to shape and grow into something real. I was tasked with identifying bottlenecks preventing the team from fully utilizing the CMS and its editor, understanding their hesitations around using AI tools, and mapping their overall workflow so we could better support them.

I spoke with several members of the research team and conducted informal, conversational user interviews. We walked through their workflows one by one, allowing them to elaborate as much as they liked on each step. They also shared input on the kinds of problems a tool like Buddy would need to solve in order for it to become part of their everyday toolkit. Documenting all of this helped our core Buddy team—consisting of the tech team, myself, and our main stakeholder—clearly define the product and the solution it aimed to deliver.

Next, close collaboration between the development team and myself was crucial to ensure Buddy struck the right balance between ambition and feasibility. I worked hand-in-hand with the AI team and front-end developers to ensure clear communication at every stage.

By the end of it, we had a tool that was ready to launch internally. But having caught the attention of the stakeholders, Buddy was now poised to soar far beyond its initial scope—toward becoming a full-fledged product of its own.