Notes on Useful Technical Work
Important work.
Unclear path.
A technical partner to clarify the problem, choose a useful path, and lead hands-on execution with your team.
Judgment, partnership, and hands-on execution by aboriginal armadillo
not a pitch deck
start with the work
Schedule a meeting01 / Field Observation
The next move is not obvious.
Important work can stall when the problem is unclear, systems do not connect, or nobody owns the technical path. We start with the people doing the work and test what actually needs to change.
question the problem, too
Symptoms
- The goal matters, but the requirements keep shifting.
- Existing tools leave people bridging the gaps.
- There are options, but no clear technical direction.
Common misread
Looks like: We need another person right now.
May be: The workflow is undefined, manual, or trapped in one person's context.
First check: Which parts repeat, which parts require judgment, and which resources already exist?
context bottleneck
this is where the pain becomes a hiring conversation
A workflow bottleneck is one place to start. Inspect before prescribing.
02 / Intervention
Clarify the path. Lead the work.
We work with your team to reframe the problem, weigh tradeoffs, and carry the technical work through. That can mean improving an existing application, connecting systems, or building software. Simplify the process first where possible. AI is one tool, not the starting assumption.
Read the field noteAn optional workflow assessmentWhen not to build an agentTechnical partnership: We lead hands-on execution. Your team brings access to the work and participates in consequential decisions.
Agent examples
Software should support professional judgment, with review before risky actions.
Use existing tools where they fit. Build where the work calls for it.
03 / Working Together
A commitment that fits the work.
Start with mutual fit, not a finished brief. Clarify what matters before choosing the work. These are questions to resolve together, not a fixed meeting sequence.
no finished brief needed
Mutual fit
Discuss context, desired future, decision participants and process, timing and broad financial fit.
Worthwhile outcomes
Clarify value, signs of progress and sensible investment. A contextual budget is not a scoped quote.
Responsible commitment
Choose clarification, bounded paid investigation, scoped implementation or no engagement. An assessment is optional.
options, not stages
assessment is optional
How we work together04 / Evidence Drawer
Records from the workshop.
Agent examples, prior work, and lessons from the workbench.



Agent examples
Agent Catalog
View catalog


Prior work
Artifacts
Open artifactsWorking lessons
Field Notes
Practical observations from building agents, workflows, and small operating systems.Read notesSelected records
- Do Not Hire Around a Broken Workflow - Why hiring pressure often points to undefined work, not only missing headcount.
- Demand Validation - A weekly demand signal brief for one offer or segment.
- The Workflow Is the Product - Why useful agent work is the system around the model call.
- AgentLabUI / Gofannon - Open-source tooling for turning agent workflows into usable applications.
- Kubeflow for Machine Learning - Production ML systems experience, from lab to deployment.
Important work. Where to start?
Tell us what matters, what is unclear, and where the work stands. No finished brief or AI project required. We can start with a conversation.
no finished brief needed
not every answer is an agent
Schedule a meetingWhat happens on the first call