Testimonials

Jonathan is an outstanding fractional AI specialist. Paired with mid level devs who have the right attitude, but are not necessarily adept in their approach and tooling practice, Jonathan can drive a team to achieve great results with a very small and carefully managed amount of time his side.

After seeing the output from the junior dev under his tutelage, and dipping in and out of their slack comms, I braced myself for a massive invoice, expecting we had racked up a huge amount of hours - however in just two hours a week, Jonathan had steered us to ship products fast, and iterate and improve really quickly. A great asset in any team looking to try and move fast in the AI space.

Jonathan has been working with Freetobook to develop an application that helps our support team respond to day-to-day user queries. The results were quickly impressive, and the tool was positively received and put to good use by our staff.  

He has helped us build, compare, and iterate on different variants of our LLM-based system. For example, we've explored router-based workflows, and trialled several approaches to incorporating retrieval-augmented context into our prompts. He has guided the continued development of the system by helping us build a user feedback system and identify key metrics that inform new variants. This data-driven approach has given us real confidence in what we can achieve with AI, with benefits extending beyond this project.

Jonathan’s collaborative approach has made him a pleasure to work with, and we would highly recommend his services.

We worked with Jonathan and Daniel from an early stage of our journey for two main purposes. Firstly, to help decide on areas of chemistry to focus on and, secondly, to help structure our AI-driven laboratory around varying levels of automation from first principles.

The engagement began with two in-depth reports to detail a global view of a complete chemistry lab set-up. They delivered beyond our expectations, including everything from the most minute details of purchasing and safe storage to an exhaustive view of the stages and equipment required to achieve closed-loop chemistry workflows. We especially appreciated receiving workflow options for manual, semi-automated, and fully automated versions of what we could achieve, including capital and operating expense estimations.

Jonathan and Daniel then worked closely with our Director of Lab & Material Innovation to support the design of a new laboratory in a new building, contributing to decisions around layout, equipment selection, and workflow constraints suitable for an AI-first approach to experimentation. We truly valued their hands-on approach and ability to identify potential pitfalls in advance, including safety perspectives. They would always try to help us understand how to balance our level of automation/complexity vs. the costs and real benefits, and helped us recognise the importance of rate-limiting steps in high-throughput design.

We are now handing over to an engineering and design firm, who will translate this foundation into a detailed technical plan ahead of construction in the coming months. Throughout the engagement, Jonathan and Daniel acted as thoughtful and reliable partners, effectively bridging AI strategy, automation, and practical laboratory operations from the outset. We look forward to continuing our work with them as we move further into the laboratory automation strategy.

I had the pleasure of working with Jonathan at Synapto, where we collaborated on building AI-driven solutions for a fast-paced fintech project. As an AI consultant, he brought a strong mix of strategic thinking and practical problem-solving that had a real impact on the direction and quality of our work.

One thing that stood out immediately was his ability to quickly identify where things weren’t working as expected and turn that into a structured approach for improvement. A great example of this was within our company enrichment pipeline, specifically the domain search step. While initial results seemed acceptable on the surface, Jonathan proposed creating a proper benchmark to measure the actual accuracy. This gave us a much clearer (and more honest) picture of the system’s performance and helped us focus our efforts where it mattered most.

From there, he played a key role in shaping the solution space. Through collaborative iteration, we moved toward a more robust multi-step approach involving domain search, evidence gathering, and candidate verification. This significantly improved the overall accuracy of the pipeline and made the system far more reliable in practice.
Beyond that, Jonathan was always thoughtful in discussions, clear in his communication, and brought a calm, structured approach to problem-solving. He has a strong ability to bridge the gap between high-level AI concepts and practical implementation, which made collaboration very effective.

I really enjoyed working with him and would gladly collaborate again. I’d highly recommend Jonathan to any team looking for someone who can bring clarity, direction, and meaningful improvements to AI-driven projects.

Jonathan joined my lab immediately following his PhD as one of our pioneering AI and Robotics specialists. Within three months, he was entrusted with the leadership of our Chemorobotics team. His first action was to establish a robust software and hardware architecture that empowered our team to develop ChemAI platforms with speed by leveraging modular, reusable building blocks. This accelerated our iteration cycles and, a decade later, several of his original libraries remain in use in daily lab operations.

Under my direct supervision, Jonathan managed a multidisciplinary team of ten engineers, PhDs, and postdocs at the then uncharted intersection of AI, Robotics, and Chemistry. His expertise in machine learning challenged the team to rethink data acquisition and successfully integrate novel concepts like active learning into our workflow.

This resulted in high-impact publications in PNAS, Angewandte Chemie, and Nature Communications, which have since gathered hundreds of citations.

What stood out most was Jonathan’s ability to understand the complexities of a chemistry environment despite his non-chemistry background. He quickly internalised the mental models and instrumentation logic required to operate at the cutting edge of the field. His ability to bridge the gap between complex robotics and chemical discovery provided a pivotal technical proof-of-concept during the formative stages of what would eventually become Chemify.

Jonathan possesses the technical depth to build the tools and the strategic vision to lead the people using them. Any organisation looking to transform their technical capabilities and accelerate innovation would be fortunate to partner with him.

I really like Jonathan’s ‘thinking out of the box’ approach which led to a much better result than tackling the problem the standard way. Overall, he is a great advisor and very knowledgeable about ML and AI. He really cares about the projects he’s supervising, goes beyond focusing on just one specific problem and considers the whole picture which makes him so great to work with.

Jonathan originally supervised my university project which is why I reached out to him when I was looking for advice on my current work project at Freetobook. Machine learning is a new area for the company, therefore I sought to get some academic consultations when solving a very complex problem. We agreed on weekly consultations which were essential in getting the project to its current stage where it is ready for user testing. I have learned a lot about all the things to consider when working on an ML based product.