Your process is manual and annoying — copy-paste, spreadsheets, and steps you redo by hand every week.
← All services
/services · Discovery
Bring me the messy version.
You don't need a perfect spec. You need someone who can help find the real problem.
This is a technical discovery session for the moment when you know something is inefficient, broken, or frustrating — but not what the solution should be. We talk it through, separate symptoms from root causes, and turn a fuzzy idea into a practical technical plan. Think of it as a problem discovery consultant who'd rather ask better questions than sell you a build you don't need.
Sound familiar?
You might be here because…
You don't have to know the answer. Recognizing the itch is enough to start.
-
-
Your tools don't talk to each other, so you're the integration holding it all together.
-
You have an idea but not a technical plan, and you're not sure what building it would even involve.
-
Your current system works, but barely — it's held together with duct tape and good intentions.
-
You know AI could help somewhere; you just don't know where it actually fits.
-
You want someone technical to explore the options with you before you commit to a build.
The output
What we turn it into
You leave with artifacts you can act on — not a vague 'let's circle back'.
Problem map
A clear picture of what's actually happening today, where the pain really lives, and what's downstream of it.
Requirements outline
The real requirements pulled out of the fuzzy ones — what a solution has to do, and what it explicitly doesn't.
Technical options
A few honest solution paths, from quick wins to bigger builds, with the trade-offs spelled out in plain terms.
Effort / risk estimate
A grounded read on what each path costs in time, complexity, and risk — no fantasy timelines.
MVP scope
The smallest version worth building first: the slice that proves the idea and earns the next step.
Recommended next step
A clear do-this-next recommendation — whether that's building, prototyping, or just changing one thing.
Process
Discovery session flow
A working conversation, not a sales call. Here's roughly how it goes.
-
01
Talk through what's happening today — the real workflow, not the org-chart version of it.
-
02
Identify the pain points and the hidden constraints: budget, tools, people, and that one system nobody dares touch.
-
03
Separate symptoms from root causes, so we fix the actual problem instead of the loudest one.
-
04
Sketch solution paths — manual fixes, automation, an internal tool, AI, or a rebuild — and weigh them out loud.
-
05
Leave with a clear next action you can act on, with or without me.
Good fits
Examples of problems I like
The kinds of messy, half-formed problems that are genuinely fun to untangle.
- Automating repetitive work that's quietly eating hours every week — the start of most automation discovery.
- Internal tools that replace a fragile spreadsheet or a manual handoff between people.
- AI-assisted workflows — finding where AI genuinely helps, with a human still in the loop.
- Infrastructure cleanup — untangling the servers, access, and backups nobody fully understands anymore.
- Security and process hardening — closing the gaps before they turn into incidents.
- Weird ideas that need technical shape — the slightly-too-big concept that needs a real plan.
Where it leads
From discovery to actually building
Software project discovery is step zero. When you're ready to build, these are the next stops.
Next step
Send me the ugly version of the problem.
No tidy spec, no polished pitch. Describe what's frustrating in plain words — technical planning for small business and weird-idea-havers alike starts exactly here.