AI knowledge systems · Content operations · Go-to-market enablement
I turn AI into real operating capacity for content, proposal and sales teams.
I've spent eleven years running content, brand and proposals at a 40-year Canadian IT firm, working alongside its managing partners and directors every day. These days most of my time goes into the AI side of that work: the agents, the knowledge bases behind them, the approval steps around them, and the training that gets people using them.
- 30 min
- from a new RFP to an 80 to 90 percent first draft, down from 1 to 2 days of setup
- 35 pages
- of RFP response drafted live, in front of leadership, by an AI agent I built
- 20-year
- contract won with a roughly 300-page response. I managed its content and assembly across a 15-person team
- 1,700+
- bids from 32 years pulled into one record that leadership can rely on for go/no-go calls
Four systems in daily use.
Each one started as a fix for something that was slowing us down. None of them publish anything on their own: a person checks every output.
Proposal Guru
Turns a new RFP into folders, branded templates and an 80 to 90 percent complete first draft in about 30 minutes. Setup used to take one to two days.
View system → Claude · SharePointSocial Media Register
Every company post is drafted here and tracked from draft to published, so approvals no longer get lost in email.
View system → Microsoft Copilot StudioCompany AI agents
Assistants for bids, marketing, Google Ads, cybersecurity and sales outreach. One drafted a 35-page RFP response live in a leadership session.
View system → Weekday digestBid Radar
A weekday morning roundup of new public-sector opportunities from 18 sources, read by the 11-person leadership and sales group.
View system →AI that leadership can trust.
Plenty of AI projects look great in a demo and then quietly stop being used. I build for the day after the demo: tools that fit how the team already works, run on the company's own knowledge and stay under human control.
- Find the bottleneckI start with the work that costs the most time or quality today, whether that's slow proposals, stuck approvals or knowledge nobody can find.
- Structure the knowledgeThen I pull scattered documents into one organised knowledge base: the approved facts, the house voice, and what the AI should never say.
- Deploy with guardrailsThe AI drafts and people decide. Gaps go to the right expert, and internal data stays internal.
- Train for adoptionHands-on sessions, a field guide and short videos, so people keep using the system long after launch.
Where I can help.
AI agents and knowledge bases
Designing agents, writing their instructions and building the knowledge they draw on, so answers sound like your company and stay within the facts.
AI content operations
Workflows that move content from idea to approved to published, with AI writing the first draft and people making the calls.
Proposal and GTM enablement
RFP responses, sales battlecards and bid intelligence, set up so a small team can compete with much larger ones.
AI adoption and training
Hands-on training, field guides and short videos that change how colleagues actually use AI from day to day.
The ELIZA Effect: Why Good AI Prompts Still Produce Bad Results
Why good AI prompts still produce bad results
I wrote and produced this two-minute video for colleagues whose good prompts were still getting bad results. It makes one point, quickly.
Produced for MYRA Systems Corp.
What I build with.
- Anthropic Claude
- Microsoft Copilot Studio
- Microsoft 365 Copilot
- ChatGPT
- Google Gemini
- SharePoint
- HubSpot
- Google Flow
- ElevenLabs
- Adobe Creative Suite
From the blog.
Practical notes on AI, content and the systems behind them. All posts
The creative work behind it.
Long before the AI work, I was a writer, designer and animator, and I still am. Here's the rest of my portfolio.