Insights & resources
Field notes from building AI deal workflows.
Practical writing for advisory and private-equity teams on governance, adoption, measurement and the boundaries of automation.
Latest
Selected writing.
Governance
What a workflow definition should contain before you build it
Trigger, inputs, outputs, approval gates, systems touched, volume assumptions and success measures. If any of these are missing, you are buying a demo, not a workflow.
Sell-side
Buyer lists are a research problem, not a database problem
Why qualified buyer universes need evidence and reasoning attached to every name – and how to keep that reasoning reviewable.
Buy-side
Target screening: the highest-frequency workflow most firms still do by hand
Standardized extraction and fit assessment against client criteria returns hours per week and makes screening decisions comparable across the team.
Measurement
Set the baseline before you launch
Turnaround time, completeness, adoption and exception rate at 90 days. Without a pre-launch baseline, every result is anecdote.
Boundaries
Where automation should stop
Client advice, valuation conclusions, external communications and fiduciary decisions stay human. The AI prepares; your professionals decide.
Adoption
Why AI pilots stall after the first quarter
No named owner, no exception handling, no maintenance plan. Operate is a discipline, not an afterthought.
Want these as a briefing for your team?
We run working sessions on AI workflow governance for advisory and investment teams.