Overnight research pass — summary (May 11-12 2026)

This page is the single readable TL;DR of the overnight pass. Read this first.

Quartz build status

  • Quartz build: CLEAN. 369 markdown files parsed, 935 emitted, 0 filtered out, 4 seconds.
  • Broken wikilinks: reduced from 11 → 6 (remaining 6 are single-reference deferred stubs: Vendr, Sierra, Kaseya, N-able, NinjaOne, concepts/outcome-pricing).

Commit chain (15 overnight commits)

From 5b1aec7 (brand rename baseline) to fad2e9d (devil’s advocate):

  1. 5d404b7 Batch 1 — 9 synthesis pages + funding corrections (Wrangle/Edra/Treeline/Aisera/Serval)
  2. 465ca1d Batch 2 — 3 synthesis pages + 8 customer-count corrections
  3. aa93221 Batch 3 — 7 synthesis pages + cross-link mining (1,884 wikilinks)
  4. 7f30813 Batch 4 — 4 synthesis pages + Anthropic-Blackstone JV entity
  5. 2f14296 Batch 5 — 2 synthesis pages + log/hot coherence
  6. 99020e1 Batch 6 — 3 fundraise-prep synthesis pages
  7. 554ea75 Batch 7 — 2 synthesis pages + Atomicwork hosting correction
  8. 24b7437 Verify Anthropic-Blackstone JV + discover parallel OpenAI Deployment Co
  9. d43e3b8 ISO 42001 cross-page consistency fix (advisor-flagged)
  10. a87ab61 Batch 8a — venture debt + token cost trends
  11. 99e8398 Batch 8b — OSS LLM viability + Atomicwork ensemble correction
  12. 414e7e6 Batch 8c — pre-IPO secondary market + exit math
  13. 3aa0b11 Batch 8 final — customer wins/churns + cumulative log/hot
  14. 522d0e5 5 high-frequency entity stubs (Microsoft, Anthropic, Google Cloud, Okta, Pax8)
  15. fad2e9d Devil’s advocate pre-mortem — bearish counterweight

What’s new (31 synthesis pages + 2 threat entities + 5 stubs)

Competitive landscape

Fundraise + capital

Engineering + technical

GTM + positioning

Brand + content + risk

Competitive-threat entities

High-frequency entity stubs

Time-sensitive action items (sorted by urgency)

  1. 🔴 Publish Aisera migration positioning WITHIN 90 DAYS. Workativ / CorePiper / Rezolve.ai already publishing “Best Aisera Alternatives 2026.”
  2. 🟡 Series A pitch deck must address 5 of 7 questions in devils-advocate-pre-mortem-2026 Section H.
  3. **🟡 Target 1B strategic exit at Moveworks/Armis 22-25× band — AI premium amortizes to par 12-18 months.
  4. 🟡 Microsoft App Accelerate application Q3 2026 (nomination beats the $100K ACR gate).
  5. 🟡 Okta OIN listing application Q3 2026 (top-3 partnership target).
  6. 🟡 Get 3-5 named customer logos before Series A pitch — Init Intelligence is currently absent from the customer-momentum scoreboard (per customer-wins-churns-2026).

Material wiki corrections

  1. Wrangle is NOT bootstrapped2M pre-seed Jan 2022 + $2M seed May 2023, Accomplice-led).
  2. Treeline FTE reframe — 70-75 = MSP rollup field staff; ~26 LinkedIn = core corporate; ~7 = AI engineering team. Holdco of 3 rolled-up MSPs, not a software startup.
  3. Edra seed = ~Dec 1 2024 close at 6M); $1.8B valuation downgraded to single-source.
  4. Atomicwork: Singapore office added (vault was missing); hosting actually Azure for AI + AWS RDS for data plane (not pure Azure); ensemble dated (Llama 2 not 3/4); Cohere CC-BY-NC = commercial-license-as-a-service NOT OSS.
  5. Aisera pre-deal 1.5B “deal price” is Series E mark not deal price (officially undisclosed).
  6. Serval $232M post-money downgraded to single-source PitchBook-via-Sacra.
  7. Moveworks ARR multiple revised 22-25× band (not 28× — SaaStr/Everest correction).
  8. ISO 42001 is now table stakes (3-way tie: Atomicwork + ServiceNow + Salesforce), not a wedge.

Highest-conviction strategic findings

  1. **1.5B + OpenAI $10B) — #1 competitive threat.
  2. Sequoia is in BOTH JVs + leads Serval + leads Edra = structurally locked out for Init Intelligence Series A.
  3. ServiceNow’s $11.7B M&A spree was 75% cybersecurity (Armis + Veza), NOT AI agents — ServiceNow positioning as AI control plane.
  4. All 4 public AI ITSM comps DOWN sharply May 2025 → May 2026 (NOW -62%, TEAM -63%, CRM -37%, FRSH -45%).
  5. Late-stage private AI bull (+165.9%) vs public AI ITSM bear = strategic-acquirer arbitrage thesis.
  6. Token cost: $2/resolution + 50K token envelope clears 75-90% GM at Sonnet 4.6 with prompt caching.
  7. MCP-native is no longer a moat (all 3 hyperscalers + ITSM incumbents have GA MCP).
  8. AI-native-buying-AI-native corridor is saturated — Init Intelligence should move upmarket (F500) or vertical-down (hospitality / credit unions / education / public sector).

Sources verified during advisor pass

The Anthropic-Blackstone JV claim was advisor-flagged as single-source-risk and re-verified across:

Verification surfaced the OpenAI Deployment Company parallel $10B JV that the original agent had under-reported.

What’s NOT pushed

  • 15 overnight commits are LOCAL only — per repo policy, no remote push without explicit user authorization.
  • Current branch: wiki/clean-lane-investor-stubs (session-start branch was wiki/competitor-cap-tables-deep-research — confirm PR target).

Remaining deferred items

  • 6 single-reference broken wikilink stubs (Vendr / Sierra / Kaseya / N-able / NinjaOne / concepts/outcome-pricing).
  • Serval NPM SecondMarket listing references 47M — flagged for verification, not yet resolved.

Where to start when you wake up

  1. Read this page (you are here).
  2. Read devils-advocate-pre-mortem-2026 — the most important page in the pass.
  3. Read customer-wins-churns-2026 — the most actionable.
  4. Read anthropic-blackstone-jv + openai-deployment-company — the new competitive threats.
  5. Decide whether to push the 15 commits to remote.
  6. Decide whether to act on the Aisera migration positioning 90-day window.