About Us
About Us
Bay Information Systems was incorporated in 2013, growing out of machine learning research at university level -- at a point when deploying an ML system required significantly more from scratch engineering than it does today. The early work centred on ML applied to embedded systems and microcontroller hardware, areas where the constraints of the environment force a level of technical rigour that carries forward into everything we do.
In 2019 we moved from independent product development into direct consulting and client engagements. That shift was deliberate: consulting funds the space to build independent tools and products between engagements, and those products keep our understanding of the full delivery pipeline -- from architecture through to market -- current and practical.
We have been building and deploying AI systems in production for over a decade. That includes the period before large language models made AI accessible as a concept, when the work required constructing the components rather than assembling them. We think that background matters: it produces a different quality of judgement about what a system actually requires versus what a vendor pitch suggests it requires.
The majority of our engagements begin before any system is built. The decisions made at the start of a project -- about what the data contains, whether the problem is well-defined, what the architecture should be -- determine whether the investment succeeds. We spend more time on this phase than most, and we produce written deliverables that give the organisation a precise, shared understanding of what they are committing to before they commit.
We bring in specialists for specific engagements where the work requires it -- engineers, designers, and domain experts with the depth the project demands. The principal remains the point of contact and accountability throughout.
We build products as well as advising on them. That means we have direct experience of the risks that sit outside the technical architecture -- the go-to-market decisions, the user adoption problems, the commercial pressures that cause projects to be descoped or abandoned. A consultant who has only ever worked inside client organisations understands the engineering risk. We also understand why something that works technically can still fail commercially, and we factor that into our recommendations.
That range -- deep technical capability combined with practical product and commercial experience -- is uncommon. Most engagements benefit from it.
AI Strategy & Engineering Consultant
Edward works with business leaders to establish whether an AI investment is sound, what it should target, and how to structure it for success. He has spent two decades building and deploying AI systems across academic research, regulated industry, and early-stage product development, with clients including the Bank of England, the Office for National Statistics, and Admiral plc, as well as startups across health, adtech, and life sciences. Recent work spans retrieval-augmented generation, knowledge graph construction, and private cloud deployment on AWS.
If you have a project you would like to discuss, or want to understand whether we might be able to help, the best first step is a conversation.