Improve everyday work
Reduce repetitive tasks, connect the right information and make everyday work easier.
- Automate recurring reports
- Cut repeated data entry
If you want to reduce repetitive work, improve a process or build something new with AI, I can help. I’m open to consulting projects or joining your team.
Where I can help
You might have an idea to test, a tool you need, or a better way of doing something.
Reduce repetitive tasks, connect the right information and make everyday work easier.
Make a tool that fits a particular need, from an internal helper to a product prototype.
Turn an early idea into a small working prototype, then try it on real tasks.
We look at what you want to improve or create, who it’s for and what would make a useful first step. Together, we agree what to test.
We try an existing tool, change a process or build a working prototype. Test it on real tasks, check the results and use feedback to improve it.
If it proves useful, we develop it in manageable stages and help the team use it. We check access, integration and human review needs before building further.
Each stage is agreed separately, with no obligation to continue beyond it. You can stop after the initial review or prototype, or keep improving and embedding what works over time.
Case Studies & Products
Generates editable compliance reports, with people retaining review and sign-off.
Project story
A useful first delivery, with people in control of the final report.
The compliance team was spending time taking supplied data and turning it into a report. Some of that work was repeatable, while commentary and validation depended on their judgement.
I worked with the team to understand where their time went and discussed what would actually help. We agreed to automate a useful part of the process first, keeping the report editable and the team responsible for its conclusions.
I produced a script that takes their data and creates an initial version of the report. The team can then add commentary, check the output and validate the finished document. Earlier work made the quantitative rules explicit and checked the generated figures against an unseen month.
I also demonstrated AI-generated commentary and running the reporting workflow as a service as proofs of concept. Those demonstrations showed what was technically possible and helped inform the phased delivery agreed with the team.
A first version does not need every possible feature to be useful. Agreeing the boundary with the team gave us something concrete to review and a sensible basis for deciding what to build next.
Supplied data becomes an editable report. The team adds commentary and validates the result.
I demonstrated AI-generated commentary. A next delivery phase could add this to the team's workflow, with people reviewing the interpretation and validating the report.
I demonstrated running the workflow as a service. Wider use depends on organisational security, data access and hosting requirements.
Illustration of the delivery approach.
Automates Jira reporting and makes gaps in delivery evidence visible.
Project story
From repetitive report production to a better understanding of the evidence underneath it.
At PokerStars, part of Flutter, a large multinational business, people were repeatedly gathering delivery data and assembling reports. That left less time to interpret what the data meant and discuss what needed attention.
I built a modular Jira reporting toolkit covering sprint reporting, strategic alignment and programme roadmaps. It separates data collection, report-specific interpretation and presentation, so another report can reuse the foundations.
I kept calculations and data checks explicit. Where a date or relationship was missing, the report made that gap visible rather than presenting an incomplete picture as a confident delivery assessment. Missing data and delivery risk needed different treatment.
Those gaps made me question whether another report was enough. I started exploring the Programme Relationship Graph: connecting delivery evidence to investigate dependencies and risk. It became a working prototype that I built and use myself.
Sprint Reporter is operational from the command line and Strategic Alignment is in use. The Programme Roadmap module is built and verified, with stakeholder sign-off on its revised layout still pending. The tools have different levels of maturity.
A shared hackathon prototype connecting evidence to missed commitments and blocked work.
Project story
Bring together the evidence that falls between workplace tools.
A meeting can contain a commitment that never becomes a ticket. A message can say something shipped while another record says it is blocked. Looking at each system separately makes these gaps easy to miss.
I contributed to a shared hackathon prototype that connects evidence from meetings, messages, documents and delivery tools. It proactively raises alerts about missed commitments, conflicting updates and blocked or stalled work, with supporting records.
Someone agrees in a meeting to deliver something, with an owner and a date. The relevant sprint closes, but no ticket records the commitment. Comparing the meeting evidence with the tracker reveals the gap that neither source could identify alone.
Detection rules are defined in code. AI can help interpret unstructured source material, but the alert points back to the records behind it. A citation opens the relevant context. When sources disagree, both are shown for a person to assess.
The finding should help someone investigate and choose a response. Human approval remains part of acting on it. Using findings to trigger further agent workflows is a future direction, not an established autonomous operation.
Independent products
Indoor cycling with trainer connectivity, workouts and virtual routes.
Project story
An indoor cycling product that joins training, connected devices and a game-like riding experience.
Apogee combines smart-trainer input, workouts and virtual routes. Workouts give the rider targets for power and cadence; routes offer courses to explore, with the option of a Pacer. Signed-out visitors can browse and try a simulated demo.
I am building the product around the complete riding journey: connect devices, choose a session, understand what to do during the ride and review the result. That means the interface and the training behaviour have to work together.
A workout and a route are different experiences, so they keep their own feedback while sharing the riding engine. Simulated demos let people explore the product before connecting hardware. Registered riding also needs clear consent and control over saved data.
I use AI to work through ideas, develop the implementation and investigate problems. The useful part is the review loop: make a bounded change, see it in the running product and check it against the intended riding experience.
The browser product is available to explore and I am working through packaging and publishing. Device connectivity, platform support and store approval need their own checks; the screenshots show a browser demonstration.

A physics playground for kids to grab, stretch and throw little creatures.
Project story
A physics playground where children discover the fun by touching and experimenting.
Jiggle Lab is a collection of little creatures that can be grabbed, thrown and stretched. Different play modes change the feel: a springy Web, a bouncy pile, drifting Zero-G creatures and a deformable Jelly, among others.
The product goal is simple: a child should be able to discover what to do without reading instructions. I build and review the interactions around what happens when someone touches the screen, and whether the response makes them want to try again.
The physics, visual feedback and sound work together. A spring network should move differently from a blob of jelly. The technical choices support those different interactions while keeping the playground responsive.
AI helps me explore ideas and build the implementation. I still need to inspect the running game: an interaction can be technically correct and feel wrong. Trying it, refining it and checking performance are part of the development work.
The browser playground is available to try and publishing is in progress. Packaging, mobile behaviour and release requirements are part of taking the prototype towards a product people can install.
Who you'll work with
I diagnose problems in how work gets done, build tools to help, and work with teams to put changes into practice. That has included delivery dashboards, automated reporting and practical AI prototypes.
I'm based in Leeds and work as Head of Enterprise Performance & PMO. My background in analytics, Agile and Professional Coaching (ICF) helps me understand how a team works and what it takes for a useful change to stick.
Tell me what you’d like to build or what’s getting in the way of your work. We can find a useful first step together.