
Hi. I'm Dee.
Creating AI systems that change how organisations work.
Agentic platforms.
LLM-powered workflows.
Internal automation that gets used.
Blending human-centred design, data, and tech to make Generative and Agentic AI adoption real.


What I do
Finding high-friction in an organisation's operations and delivering AI systems to eliminate them. Product lens, end-to-end. From the messy first conversation about the problem to something running in production.
Human workarounds hold most organisations together. That's not a criticism, it's a design opportunity. When you design for agents from the start, the people can finally do work that matters.
What I've built
MY IMPACT
01 / Delivery Intelligence
Ongoing innovation
Designing, experimentingand delivering internal AI automation to eliminate the highest-friction points in how we run AI delivery programs at Kablamo.
Backlog Agent: pulls requirements directly from Confluence, generates structured epics, links user stories, and tickets with acceptance criteria. Eliminated hours of manual backlog planning per sprint.
Action Agent: ingests live transcripts, extracts decisions and owners, drafts follow-up comms, logs tasks automatically. Running live. Zero manual minutes-taking.
Project Dashboard: live data from Jira, Asana, Confluence RAID boards and Slack signals. Claude generates weekly narrative status reports with auto-derived RAG status, blocker summaries, and decision flags. In build.

01 / Delivery Intelligence
Ongoing innovation
Designing, experimentingand delivering internal AI automation to eliminate the highest-friction points in how we run AI delivery programs at Kablamo.
Backlog Agent: pulls requirements directly from Confluence, generates structured epics, links user stories, and tickets with acceptance criteria. Eliminated hours of manual backlog planning per sprint.
Action Agent: ingests live transcripts, extracts decisions and owners, drafts follow-up comms, logs tasks automatically. Running live. Zero manual minutes-taking.
Project Dashboard: live data from Jira, Asana, Confluence RAID boards and Slack signals. Claude generates weekly narrative status reports with auto-derived RAG status, blocker summaries, and decision flags. In build.
02 / LLM-Powered Archive & Retrieval System
4 months, Principal
Led the design and delivery of an AI accelerator program for a major national media organisation, embedding generative AI into their digital archive platform. Co-designed with archivists, technologists and executives to build adoption workflows, QA and validation loops, and shared success metrics. Outcome: 92% reduction in archive search time.
Delivered the AI metadata transformation initiative, integrating LLMs into archival workflows to enrich and validate metadata.
Built QA checks and human validation workflows to catch errors early and ensure accuracy, accountability, and compliance.
Designed AI adoption strategy, aligning executives, archivists and engineers around shared success metrics.

02 / LLM-Powered Archive & Retrieval System
4 months, Principal
Led the design and delivery of an AI accelerator program for a major national media organisation, embedding generative AI into their digital archive platform. Co-designed with archivists, technologists and executives to build adoption workflows, QA and validation loops, and shared success metrics. Outcome: 92% reduction in archive search time.
Delivered the AI metadata transformation initiative, integrating LLMs into archival workflows to enrich and validate metadata.
Built QA checks and human validation workflows to catch errors early and ensure accuracy, accountability, and compliance.
Designed AI adoption strategy, aligning executives, archivists and engineers around shared success metrics.
03 / Enterprise Agentic Platform
Ongoing, Principal and Client Lead
Leading product strategy and delivery for a 7-agent AI system transforming Sales and Ad Operations workflows. Owned the PRDs, system design principles, and operating model. Ran discovery with business and exec stakeholders, defined what each agent needed to do and how they'd be evaluated, and worked alongside the lead AI engineer to translate product vision into production. Responsible for phased rollout planning and stakeholder alignment across the program.
Owning the PRDs, system design principles and operating model: defining what each agent does, how outputs are evaluated, and where humans stay in the loop.
Running discovery and stakeholder alignment across business, exec and Ad Operations teams, translating messy business problems into a structured product vision the engineering team can align with and build to.
Working alongside the lead AI engineer and cross functional teams to take the product from concept to phased rollout through sequencing, quality and reporting across the program.

03 / Enterprise Agentic Platform
Ongoing, Principal and Client Lead
Leading product strategy and delivery for a 7-agent AI system transforming Sales and Ad Operations workflows. Owned the PRDs, system design principles, and operating model. Ran discovery with business and exec stakeholders, defined what each agent needed to do and how they'd be evaluated, and worked alongside the lead AI engineer to translate product vision into production. Responsible for phased rollout planning and stakeholder alignment across the program.
Owning the PRDs, system design principles and operating model: defining what each agent does, how outputs are evaluated, and where humans stay in the loop.
Running discovery and stakeholder alignment across business, exec and Ad Operations teams, translating messy business problems into a structured product vision the engineering team can align with and build to.
Working alongside the lead AI engineer and cross functional teams to take the product from concept to phased rollout through sequencing, quality and reporting across the program.
What I've built
TOOLS & FRAMEWORKS
My Approach
Beliefs behind my work
01. Governance is design.
Compliance and trust aren’t add-ons; they’re built into the workflows, data models, and systems from the start. Effective governance is making it usable, adoptable, and safe at scale.
02. Strategy means nothing without delivery.
Big ideas only matter if they land. That’s why I work at the intersection of strategy, design, and execution.
03. Design for the agent, not the workaround.
When you design for agents from the start, the people can finally do work that actually matters.
04. Outcomes matter more than outputs.
The measure of success isn’t the launch of a system, but the value it creates for customers, employees, and organisations.
01. Governance is design.
Compliance and trust aren’t add-ons; they’re built into the workflows, data models, and systems from the start. Effective governance is making it usable, adoptable, and safe at scale.
02. Strategy means nothing without delivery.
Big ideas only matter if they land. That’s why I work at the intersection of strategy, design, and execution.
03. Design for the agent, not the workaround.
When you design for agents from the start, the people can finally do work that actually matters.
04. Outcomes matter more than outputs.
The measure of success isn’t the launch of a system, but the value it creates for customers, employees, and organisations.



