Directions Brain

vertical-use-case-led-brain

5 min read

Vertical Use-Case-Led Brain

Definition

The strategic repositioning of the Brain (context-os-brain) from a generic horizontal context layer into a vertical, use-case-led product: pick the workflows that genuinely require deep cross-team context, build the agents that win those workflows, and let the Brain be the secret sauce / moat sitting underneath. Articulated most clearly by guy-barkat in 2026-05-27-directions-vertical-pivot-and-prediction-markets and reinforced by saar-arbel.

Key points

  • Ceiling argument. A generic “wiki-like Brain” cannot be sold for $1M ARR — buyers will not pay enterprise prices for “store our files better.” A vertical, use-case-led product can attribute concrete output (X% faster delivery, replaced agency contract, blocked compliance regression) — that unlocks six-figure-plus deals.
  • Reverse the design direction. Today the team starts from “structure the data” and asks “what can agents do on top.” Guy’s argument: invert. Start from the highest-value agentic workflow you can credibly own, and design the Brain shape backward from that workflow’s context requirements.
  • The “Wonderful for X” pattern is the proof point. notch (Wonderful for regulated industries / call centers — where maor-naksh is moving), alta (Wonderful for revenue cycle, by Stav and team) are real, funded executions. The pattern works.
  • Champion + ROI matter. A vertical product has a clear champion (the buyer who personally wins from the workflow improvement). A horizontal Brain has no obvious champion — it’s “useful to everyone but owned by no one.”
  • Don’t jump the vertical pick. Guy explicitly pushed back against jumping to construction or healthcare without first checking the team’s home turf (hi-tech enterprise — etoro / monday-shaped) for un-served use cases. The home turf may be the right vertical because the team understands it.
  • Compatibility with the agent-native thesis. This framing extends agent-native-go-to-market rather than replacing it: the Brain is still queryable by agents day-1, but the agents are yours, not anyone’s, and you sell on their workflow output. Pricing model can ride the agent’s token usage (Cursor-like 20% margin on tokens) instead of seat-based wiki licensing.

Evidence

  • 2026-05-27-directions-vertical-pivot-and-prediction-markets — Guy’s central argument; Saar’s ari-leshno / eye-clinic anecdote as a concrete vertical candidate; Nizan’s concession that the GTM is stronger when value is use-case-led.
  • 2026-05-19-saar-leshno-glaucoma-data-tracking — the source of the eye-clinic anecdote: a concrete healthcare wedge (ari-leshno’s glaucoma visual-field tracking pitch) plus the underlying generalizable pattern, clinical-data-portability. Important caveat: the wedge as pitched is closer to a healthcare SaaS than a Brain-shaped product, and the open question for this topic is whether the Brain layer adds real value on top of CRUD-with-OCR.
  • 2026-05-06-brain-os-strategy-brainstorm — earlier framing that “the Brain needs a sharper next-2-steps story beyond Operating System” — this topic is the team’s first concrete answer to that prior open question.
  • 2026-05-01-entree-capital-enterprise-ai-spend-map — external VC validation of the vertical-AI thesis: Bessemer reports vertical-AI portfolio +400% YoY at ~65% gross margin; Scale VP & NEA frame the addressable market as 450B SaaS (see vertical-ai-tam); first vertical AI IPOs expected 2027–28 in legal, healthcare, and financial services. Embedded AI in existing SaaS forecast to lose share (35% of single-purpose SaaS replaced by AI agents by 2030, Gartner) — reinforces the “don’t build inside Monday’s seat structure” cut of this thesis.
  • 2026-03-06-alonhuri-linkedin-saas-is-dead-ai-native-agency — most aggressive cut of this thesis from alon-huri (team8): be the vertical service provider, not the software vendor. Names 15 regulated professions where one licensed human + AI back-office can replace a 10–50-person firm. Different revenue/exit shape than “Wonderful for X” — pulls the discussion toward the Full-Stack AI direction. See full-stack-ai-vertical-services for the synthesis. Skepticism caveat — Team8 portfolio alignment.
  • 2026-01-30-alonhuri-linkedin-full-stack-ai — companion to the above: theoretical framing of Full-Stack AI as services-company-that-becomes-AI-native, with proprietary data loop + distribution engine as the two structural moats; 5x–10x bar required to justify replacing an incumbent industry. Anchors full-stack-ai-vertical-services.
  • 2026-05-31-directions-connect-businesses-and-brain-scaling — Saar reinforced that vertical determines the Brain’s data model (a construction-Brain has no companies/entities — it has patients-equivalents, treatments, sites). Guy reached the “difficulty is the moat” reframe — narrowing scope to one vertical’s data shape is what makes the agent harness tractable for a startup while incumbents can’t replicate it. NOTE: this meeting also opened a non-Brain third direction (connect-businesses-multi-vertical-api), so vertical-Brain is no longer the team’s only active lens.
  • 2026-06-02-directions-account-management-vertical — Convergence: Account Management / Customer Success is the team’s primary vertical candidate. Prior Brain analysis showed 79% fit. saar-arbel demoed a per-AM Brain built from daniela’s last ~100 call transcripts. nizan-shifman’s VC framing (nir-goldstein): if you can grow the book of business AMs manage from 20%→70%, that is the billion-dollar proof point. Founder-market fit: Saar + Nizan built Monday CRM to $140M, mostly on AM use cases.

Open questions

  • Which vertical first? Working answer as of 2026-06-02: Account Management / Customer Success. Still to validate via the daniela pilot and practitioner feedback (ofer-mixpanel, Gal from Dave-AI startup).
  • Is the right unit “Wonderful for vertical X” (one product per vertical) or a Brain platform with vertical packs on top? The former is simpler; the latter scales better.
  • How much of the monday-agents-month catalog actually requires a Brain vs being thin Cursor wrappers? Mining required.
  • Pricing: per-seat licensing of the Brain itself, or token-margin on agent usage, or hybrid? Not resolved.
  • Does this reposition sacrifice the “first-to-market on horizontal context-OS” claim, or strengthen it because the wedge is sharper?

Graph View

Backlinks