Directions Brain

Directions at Guy's House - Revenue-Cycle Brain and Construction POC

3 min read

Directions at Guy’s House - Revenue-Cycle Brain and Construction POC

Source: raw/meetings/Directions (2026-06-13 20.00) At Guy׳s house.txt

Summary

In-person directions session at guy-barkat’s house. The team compared three active wedges: a new revenue-cycle-brain thesis, the existing account-management-vertical, and a construction Brain POC path grounded in Saar’s warm construction contact / almogim-exec-visibility-poc. The strategic through-line was sharper than prior sessions: generic context storage is not enough; the product has to create concrete business output, and the Brain is only valuable insofar as it powers that output.

Key takeaways

  • Generic context-AI pitches are not enough. Employee interviews, workflow recording, and model-agnostic memory sound useful, but the team kept returning to the same test: what business result does the customer buy?
  • wonderful remains the clearest operating pattern. Guy reported a major discount-bank deployment and the team debated its FDE-heavy reality: painful implementation work, but real customer value, expansion potential, and a platform being built underneath the service motion.
  • A new revenue-cycle-brain thesis emerged: connect marketing, sales, product, support, and retention data into one closed loop so AI can optimize the full customer lifecycle rather than local marketing metrics.
  • The revenue-cycle wedge is strategically large but operationally dangerous. Every customer has different data models, identity joins, delays, tools, and semantics; the team was explicit that this could collapse into bespoke data-engineering unless the first wedge is narrow and value-first.
  • account-management-vertical stayed attractive because it is more bounded and less crowded than marketing automation. The job is to grow each AM’s book of business, automate back-office relationship work, and preserve human personal-touch moments for the accounts that matter.
  • Construction stayed live as a design-partner path. Saar’s construction contact wants an executive cockpit, protocol memory, daily push updates, and a WhatsApp-style company brain; this could become a paid POC and force the team to learn the domain.
  • The team aligned on a faster validation rhythm: choose a wedge, build a demo/POC quickly, show it to real operators, and use the feedback loop to decide whether to double down or move on.

Decisions

  • No final vertical decision. The team explicitly left the choice between revenue-cycle, account-management, and construction open and planned to revisit after more thinking / validation.

Action items

  • Team → Choose the wedge to work on this week and finish a demo or POC that can be shown to real operators by the end of the week.
  • saar-arbel → Continue warming the construction design-partner path and test whether the exec-cockpit / company-brain demo triggers real willingness to pay.
  • Team → Validate the account-management-vertical with operator conversations, including the ofer-mixpanel path if available.
  • Team → Pressure-test the revenue-cycle-brain wedge by defining the smallest data-source set and buyer that can produce Day-1 value.

Open questions

  • What is the Day-1 product for revenue-cycle-brain: campaign optimization, sales-call insight mining, landing-page changes, AM/support interventions, or something narrower?
  • Who buys the revenue-cycle product first: CMO, CRO, CEO, growth lead, or a new AI-transformation owner?
  • Does account-management-vertical beat revenue-cycle because the implementation surface is bounded and the buyer pain is clearer?
  • Can the construction POC become a VC-grade vertical, or is it more likely a high-touch/bootstrap software-services path?
  • What part of the Brain must be productized first: passive context ingestion, active clarification/human-in-the-loop, or the agents that take action on the context?

Graph View

Backlinks