How I work
Assess. Design. Build. Adopt.
Not every project needs all four. Some clients want a second set of eyes on their systems. Others know what they need and want it built right. Every step ends in something you can look at, not a document about it.
Who this is for
- Size
- Organizations of 10 to 200 people. Big enough that the spreadsheet is a problem, small enough that nobody owns fixing it.
- Stage
- You've outgrown spreadsheets and email, and you don't have a data or product team of your own. Leadership is smart but not technical.
- Where I've worked
- Insurance brokerages, financial advisory firms, government and education programs. The situation matters more than the industry.
Step 01
Assess
Figure out what you have
I dig into your systems, your workflows, and the workarounds nobody wrote down. Conversations with the people doing the work, not just the people managing it. You walk away with a clear picture of what to keep, what to replace, and what to build.
- A review of your systems, workflows, and data. What you have, what's broken, what's duplicated
- Workflow maps that show where the time goes
- Vendor and platform evaluation when you're deciding what to buy
- A prioritized roadmap, not a 50-page report
- Where AI could actually help, and where it would be a distraction
- Usually a rough prototype of the first thing worth building
Step 02
Design
Decide how it should work
Before anything gets built, you see it. I map how your team actually works, design the tool around that, and give you a working prototype you can click through. You'll argue with the thing before a line of production code exists.
- Solution architecture that shows how the pieces should fit together
- UI/UX design for internal tools and client-facing apps
- Journey maps grounded in how your team actually works
- Working prototypes you can click through and react to
Step 03
Build
Build the fix
Custom applications, integrations between the systems you already pay for, automated reports, dashboards with the "so what" built in, and AI where it earns its keep. I pick tools your team can maintain after I'm gone. If a no-code option does the job, I'll say so.
- Custom web applications built around how your team works
- Integrations between the systems you already pay for. Anything with an API, and years inside Salesforce, Pipedrive, and HubSpot
- Workflow automation that removes manual steps and reconciliation
- Automated reports and dashboards that tell a story, not just display numbers
- AI workflows with a person in the loop. Reading documents into structured data, drafting the first pass, flagging what needs a human
Built with
- Apps
- R Shiny, Vue, JavaScript, CSS
- Reports and dashboards
- Quarto, Observable, Power BI, Looker
- Integrations
- Anything with an API. Deep CRM experience in Salesforce, Pipedrive, and HubSpot
- AI
- Large language models wired into the workflow, never left unsupervised
Step 04
Adopt
Make it stick
A tool nobody uses is an expensive way to change nothing. This is the step everyone else skips. Design is what makes a tool something people open twice. Rollout, training, and docs are what make it stick after that.
- Rollout planning and training
- Plain-language documentation
- Support through the first real cycles, whether that's month-end, quarter-end, or enrollment season
- A handoff so you're not dependent on me
Let's talk
Got a data problem? Let's talk.
Tell me what's going on. No commitment, no forms that take 10 minutes.
Rather talk it through? Pick a time. Not ready to talk? Take the two-minute quiz.