Legal AI  ·  Research Memo

Ivo — the F500 contract review leader

Best logo velocity in the in-house AI legal assistant segment — but three of its own customers, unprompted, named Ironclad / DocuSign / Microsoft Copilot as the system that displaces it.

May 2026

1

The legal AI stack has five layers — Ivo sits in the in-house cognitive cell as the F500-credentialed targeted review tool

Point Solutions

Narrow tasks: redlining, drafting, clause libraries

Spellbook, Draftwise, Wordsmith

AI Assistants (cognitive) ← Ivo

Chat, research, drafting, review; BigLaw or in-house

Harvey, Legora · GC AI, Ivo, Ruli AI

CLMs / SoR

System-of-record for contracts, now absorbing AI

Ironclad, Luminance, Conga

Orchestration / Intake

Ticketing, triage, agentic workflows for in-house

Sandstone, Arca, Checkbox

Horizontal LLMs

Commoditization threat from below

Claude Cowork, Copilot, ChatGPT

The central question in every legal AI memo

Can the AI-assistant wrapper survive as frontier LLMs improve, or does value migrate up (CLMs absorb intelligence — Ironclad Jurist[22]) or down (LLM platforms absorb workflows — Claude Cowork)? Ivo’s bet: Word-native review quality + Intelligence repository builds enough data gravity before the window closes.

Source: Altis Harvey-v-Legora, GC AI, Sandstone, Wordsmith memos (Apr 2026); expert calls (N=20+ tangential)
2

F500 logo velocity is real — but 3 of 3 Ivo customers, unprompted, named Ironclad / DocuSign / Copilot for Legal as the displacement path. The bundling clock is the thesis.

Bull — F500 logos + Word-native wedge + earliest repository expansion
  • F500 logo velocity leads the segment: IBM, Uber, Shopify, Atlassian, Reddit, Canva, Pinterest, CDW. 250% F500 growth YoY; denser roster than GC AI, Sandstone, or Ruli combined.[6]
  • Word plug-in + playbook-mirroring is the decisive purchase frame: All 3 May 2026 customer voices independently disqualified rivals on this. Consumer Internet voice ruled out Ironclad Jurist for lacking a Word plug-in. Min-Kyu Jung shipped the first LLM-powered Word redlining add-in Dec 2022.[11]
  • NPS 9-10 across all 3 customer voices: Ceiling is buyer adoption, not product gap. “It’s software that makes you smile” (Data Center AGC). Practitioner ranking: “Ivo above GC AI.” 85% trial win rate.[8]
  • Pricing well below buyer-elasticity ceiling: Mid-market 6-license deal: $40–60K/yr; same buyer named $100K as “still worth it”; headcount alternative $150–200K. Renewal headroom is real.
  • Repository landing as Evisort replacement: Ivo Intelligence (Jun 2025)[7] flagged by mid-market GC as Evisort displacer. ~17x implied ARR multiple vs. ~55x GC AI.[1]
Bear — Low switching costs, small SAM, three-way category compression
  • CLM-bundling / Copilot displacement is the load-bearing bear case — named by 3 of 3 customer voices, unprompted: Data Center AGC: “Wouldn’t be shocked if Ironclad acquires Ivo’s functionality.” Consumer Internet Commercial Counsel: “Cannibalism with the bigger players of smaller tools.” Mid-Market Tech CLO: “Maybe a full suite is the way to go.” Active customer-articulated path, not theoretical.
  • Switching costs are behavioral, not architectural: All 3 voices agree data/integration lock-in is effectively zero. “Very easy to offload after one year — just our playbooks. Doesn’t plug into any of our major systems.” Playbook-curation creates moderate stickiness; if a bundled equivalent ships, friction collapses to “we’d move at renewal.” (Nuances corpus §3.)
  • CLMs + Copilot moving up-stack: Ironclad Jurist,[22] Luminance AI-native, Conga + LinkSquares + DocuSign/Lexion all embedding AI review. Microsoft Copilot for Legal is the unmodeled fourth front — owns Word distribution.
  • LLM commoditization accelerating: Claude Cowork legal plug-in shipped Mar 2026. Where enterprise Claude/GPT/Copilot licenses exist, marginal cost ≈ $0.
  • In-house SAM is $0.5–1B globally: Concentrated US ~$450–550M (Altis Wordsmith memo). Small pool, 8+ contenders — cluster cannot support multiple venture-scale outcomes without expanding into adjacent workflows.
  • Adoption gates per-account expansion: All 3 voices flagged playbook-curation + change-management as the binding constraint. Usage saturation lags purchase 6–12 months even with motivated buyers; seat-count expands slowly.
  • ARR is inference, not disclosure: ~$15–25M implied. No public NRR or Intelligence attach rate. F500-scale pricing has only one mid-market datapoint.

“Wouldn’t be shocked if Ironclad acquires Ivo’s functionality. We’d go, ‘Cool, because it integrates with Ironclad.’”

— Director / AGC | Data Center Infrastructure (Ivo customer)

“It’s a software that makes you smile, which is a weird thing to say. I’d give it 10/10.”

— Same voice, on product NPS

“Ivo was hungry to meet our needs. As you start getting into the bigger platforms, the prices just go crazy.”

— Chief Legal & Procurement | Mid-Market Tech
Source: 4 direct Ivo-customer expert calls (May 2026, 3 distinct voices) + 20+ tangential calls; Legally Disrupted podcast; Altis Legal AI reference memos (Wordsmith, GC AI, Sandstone, Harvey-v-Legora)
3

Contents

01
Company
Ivo's founding story, funding stack, product architecture, and the Word-native wedge that separates it from peers
02
Competitive
Landscape positioning, implied multiples, and the three-way compression dynamics squeezing the in-house AI assistant segment
03
Risks & Signals
F500 customer signals, expert sentiment, and the six signals that determine category leader vs. acquisition target
4

Founded 2021 in New Zealand as Latch[9]; Blackbird-led $55M Series B at $355M post in Jan 2026[10]

Founders & Team

  • Min-Kyu Jung (CEO) — ex-corporate lawyer at Bell Gully (Auckland, NZ); self-taught coder in 2021.[14][12] Built first LLM Word redlining add-in Dec 2022.[13]
  • Jacob Duligall (CTO) — ex-senior SWE at Xero (Wellington, NZ).[20]
  • ~80 employees → targeting 180 by EOY 2026 (London + NY offices opening)[6]
  • HQ: San Francisco (relocated from Auckland); 5-days-in-person

Funding stack ($77M total)

RoundDateAmtLead
Pre-seed~2023Daniel Gross[19]
SeedApr 2024$4.8MFika + Uncork[4]
Series AFeb 2025$16MCostanoa[5]
Series BJan 2026$55MBlackbird @ $355M post[1]

Growth metrics (Mar 2026, self-reported)

6x
ARR YoY
134%
Customer growth
250%
F500 growth
85%
Trial win rate
RepVue (third-party): Top 5% of sales orgs globally; #3 worldwide inbound lead flow; #8 best VC-backed sales org

ARR Reverse-Engineering

Ivo does not disclose absolute ARR.

  • Anchor: $355M post at Series B[1]
  • Multiple: 15–25x (6x-growth vertical SaaS)
  • Implied: ~$15–25M ARR
  • Cross-check: Artificial Lawyer “5Xs ARR”[3] → internally consistent

Caveat: Inference, not disclosure. Verify in diligence.

Source: NZ Herald, Capital Brief, Artificial Lawyer, BusinessWire, GlobeNewswire, Crunchbase, LinkedIn, LawFuel
5

Ivo reviews contracts inside Microsoft Word for in-house legal teams — it redlines like a senior attorney and builds a searchable clause repository

Who buys it

In-house general counsel, legal ops leads, and in-house attorneys at mid-market and enterprise companies. Explicitly not law firms, not solo practitioners. Named customers include IBM, Uber, Shopify, Atlassian, Reddit, Canva, Pinterest, CDW, Pipedrive, WeightWatchers, Eventbrite, Quora, Geotab, Fonterra, WordPress, and BCBS Kansas City.[6]

What it does

Reads an inbound contract inside Word, applies the company's playbook (approved clauses, red-line positions, risk thresholds), and produces tracked-changes redlines that mirror how a senior attorney would mark up the document — consistent terminology, minimal edits, defensible reasoning. Then stores every reviewed contract in a searchable repository and benchmarks clauses against internal + external market data.

Why it matters (13 published case studies)

  • Accordion: $132K saved in 6 months, 349 contracts reviewed, 221h recaptured[25]
  • Vantage Data Centers: 85-person legal team; all 3 Ivo products deployed.[26] Expert-call corroborated 10/10 NPS — but the same Vantage voice expects Ironclad / Copilot to displace Ivo within 2-3 years. Case study is not a retention commitment.
  • Quora: evaluated 13 AI review tools over 6 months; selected Ivo[18]
  • CNG Holdings: 20–40h/week saved; Word plugin deployed in 5 minutes[27]
  • NDA turnaround: 4 days → 1 day (customer-tracked KPI).[11] Geotab: 1h → 15min.[17]

How it works & pricing

Word add-in (Win + macOS). Three products: Review, Intelligence, Search Agent. CLM-agnostic. Mid-market 6-license deal: $40–60K/yr; buyer’s elasticity ceiling ~$100K/yr; alternative cost (headcount + external counsel) ~$150–200K/yr. F500-scale pricing: undisclosed.

Source: ivo.ai product pages; case studies; GlobeNewswire (Feb 2025, Mar 2026); Altis GC AI memo (Apr 2026)
6

400+ model calls per review + Word-native execution + Intelligence repository = platform-expansion path earlier than any peer

Product modules

  • Ivo Review — MS Word add-in for contract redlining. Custom playbooks. Claims 97% accuracy on CUAD benchmark (Contract Understanding Atticus Dataset).[15]
  • Ivo Intelligence (Jun 2025) — Contract repository analytics. Extracts clauses with no setup/tagging. Benchmarks vs. internal + external data.[16][7][8]
  • Ivo Search Agent (Feb 2025) — Cross-portfolio search (Box, SharePoint, CLMs). No manual metadata tagging.[5]
  • Ivo Assistant — Plain-language prompt-based edits inside Word.

Architecture & differentiation

  • 400+ model calls per review — agentic chaining; multi-model architecture
  • Dynamic model routing — small specialized models for simple tasks, frontier for complex. Claims 10x faster, 25x cheaper than frontier-only.[28] Self-hosting for data residency.
  • CLM 2.0 positioning — Intelligence extracts insights across millions of agreements without metadata tagging, maps amendment chains. Positions vs. CLMs, not alongside them.
  • Known gap: no drafting — Ivo reviews and redlines but cannot draft from scratch (confirmed by Spellbook comparison). Real product limitation.

“Most AI tools are like a smart junior lawyer who joined your company five minutes ago and have no context whatsoever. They don't know that you sell shoes. They don't know your preferred negotiating positions or where the risk tolerances are.”

— Min-Kyu Jung, CEO | Ivo — Legally Disrupted Ep. 39 (the “onboarding problem”)

“We have a customer who told us they've gone from four days on average to turn around an NDA down to one day, and they know this because they use their CLM to track turnaround times. What does it mean if your sales team can close deals four days earlier?”

— Min-Kyu Jung, CEO | Ivo — Legally Disrupted Ep. 39
Source: ivo.ai product pages; Legally Disrupted podcast; tangential expert calls
7

Ivo is the ARR + logo leader of the in-house AI assistant cluster — and priced more reasonably than GC AI on implied multiples

CompanyFoundedARRValuationRaisedFTEPositioning
Ivo2021 (NZ)~$15–25M impl.[3]$355M[1]$77M~80Word-native review + Intelligence; F500 logos
GC AI2023~$10M[23]~$555M$72M~70In-house CEO-led PLG; 1-10 lawyer teams
Sandstone2025Undisclosed[24]TBD$10M~20Intake + orchestration layer
Spellbook2018Undisc.Undisc.~$23M~100Drafting-first; solo/small-firm strength
Ruli AI2024Undisc.TBD~$20M<50“AI GC” positioning; NYC-based
Pincites2021Undisc.Undisc.Undisc.<20Boutique Word plugin; practitioner favorite
Harvey (in-house)2022~$190M+[21]~$11B~$1.2B~1,100Law-firm platform moving down-market

Pricing: Category consensus ~$450–550/user/month (up from $150–180 in 2023). All tools priced comparably — Harvey, GC AI, Ivo, Spellbook same tier. Implied ARR multiples: Ivo ~17x; GC AI ~55x; Harvey ~58x. Ivo is the most reasonably priced name in the segment.

Defensibility hot spots: Leads on Word plug-in + playbook-mirroring + F500 logos. Loses on data gravity vs. Ironclad, distribution vs. Harvey, founder-channel vs. GC AI. Architectural switching cost is effectively zero per customer voices — bundling threat (Ironclad / DocuSign / Microsoft Copilot for Legal) was named unprompted by 3 of 3 May 2026 Ivo customers as the displacement path.

Source: Altis Harvey-v-Legora, GC AI, Sandstone memos (Apr 2026); NZ Herald, Artificial Lawyer, Crunchbase
8

Three independent Ivo customers, unprompted, named Ironclad / DocuSign / Microsoft Copilot for Legal as the displacement path — the bundling threat is no longer theoretical

LLMs flattening from above

Claude Cowork legal plugin shipped Mar 2026. Marginal cost ≈ $0 where enterprise Claude, GPT, or Gemini licenses exist.

Most in-house teams report material capability loss switching to a general-purpose LLM today — but the gap is closing each release.

Microsoft Copilot owns Word distribution; one legal-specific update could re-level the playing field.

CLM / Copilot bundling from above & below

3 of 3 May 2026 Ivo customers named Ironclad-or-equivalent CLM, DocuSign, or Microsoft Copilot for Legal as the system that displaces Ivo — unprompted, in different interviews.

Ironclad Jurist — agentic reviewer launched inside CLM base; owns data gravity.[22] Luminance, Conga, LinkSquares, DocuSign/Lexion all embedding AI review into the system of record.

Microsoft Copilot for Legal owns Word distribution — the unmodeled fourth front. Architectural switching cost for Ivo is effectively zero per customer voices.

Harvey descending from law firms

Harvey at ~$190M ARR, ~$11B val, moving into mid-market + in-house.[21]

Harvey insiders cite Ivo's Intelligence roadmap as encroachment target.

Legora (~$100M ARR) is the other platform descending into US + in-house.

“I wouldn’t be shocked if Ironclad or some other large CLM company acquires Ivo’s functionality to incorporate into a broader contract management software... I think eventually there’s going to be a lot of cannibalism with the bigger players of smaller tools. Maybe a full suite is the way to go.”

— Composite of 3 independent May 2026 Ivo customer voices (Data Center Infrastructure AGC; Consumer Internet Commercial Counsel; Mid-Market Tech Chief Legal & Procurement)
Source: 4 direct Ivo-customer expert calls (May 2026, 3 distinct voices); Altis Harvey-v-Legora, GC AI, Wordsmith memos; Ironclad/Luminance product pages
9

F500 roster is denser than in-segment peers combined and NPS lands 9-10 across all 3 May 2026 customer voices — but the same voices expect Ivo to be replaced by an Ironclad / Copilot bundle within 2-3 years

IBM
Uber
Shopify
Atlassian
Reddit
Canva
Pinterest
CDW
Pipedrive
WeightWatchers
Eventbrite
Quora
Geotab
Fonterra
WordPress
BCBS KC

“Nine [out of 10]. This doesn’t have anything to do with actually the product. It has to do with the customer’s adoption... it’s more of a human issue where there’s change management that is involved.”

— Chief Legal & Procurement Officer | Mid-Market Tech (current Ivo customer)

“Our biggest criteria was making sure there was an AI Word plug-in... [Ironclad Jurist] would have to review documents in a web browser-based document. They don’t have a Microsoft Word plug-in.”

— Commercial Counsel | Large Consumer Internet Co. (current Ivo customer)

“It’s very easy to offload after one year. It’s just our playbooks. It’s nothing highly sensitive that we’re sharing with Ivo. It doesn’t plug into any of our major systems. It’s just a simple Word plug-in.”

— Commercial Counsel | Large Consumer Internet Co.

“If I had to rank their Word plug-in, Ivo is still above GC AI. Pincites, on the Word plug-in, is still the best.”

— In-House Counsel | Mid-Market SaaS (tangential, prior corpus)

“[Ivo] recently added a repository feature which we’re considering looking at and potentially getting rid of Evisort.”

— General Counsel | Mid-Market (tangential, prior corpus)

“Of those four, I would invest in all four but Harvey. Ivo has a great reputation and great followers.” [Ivo 7/10; GC AI 6; Harvey 4]

— GC + COO | 1,100-Person Mortgage/Fintech (tangential, prior corpus)
Source: 4 direct Ivo-customer expert calls (May 2026, 3 distinct voices); 20+ tangential calls; GlobeNewswire (Mar 2026)
10

Three load-bearing holes from the May 2026 customer pass — plus three category-leadership signals. Watch the bundling clock first.

  1. Ironclad / DocuSign / Microsoft Copilot for Legal roadmap timing [medium-confidence hole — 3 customer voices converging makes this active, not theoretical]. Fill from: direct call with Ironclad/DocuSign product leader OR Microsoft Copilot for Legal PM. If resolved: 12-18 months out = Ivo wins on inertia; 24-36 months out = Ivo’s window to lock in F500 ARR widens. Blocked by: vendor-roadmap calls are scarce on expert networks.
  2. F500-scale pricing [low-confidence hole — one mid-market $40–60K datapoint doesn’t generalize to 20–100+ seat F500 accounts]. Fill from: 5+ pricing data points across F500 buyers. If resolved: tests whether implied ~$10K/seat/yr unit pricing holds at scale or compresses. Blocked by: F500 buyers harder to source; pricing often NDA-bound.
  3. Procurement-only buyer persona [low-confidence hole — one joint legal+procurement voice said procurement is not an active user; one account had no procurement function at all]. Fill from: 5-10 calls with stand-alone procurement leaders at Ivo accounts. If resolved: if procurement drives some purchases, pricing-power story changes (procurement-led = pricier-but-shorter; legal-led = smaller-but-stickier). Blocked by: no current procurement-only contacts in the network for Ivo accounts.
  4. Net Revenue Retention. In a category where architectural switching cost is effectively zero per customer voices, NRR is the only durable moat signal. Below 120% reframes the standalone thesis. Watch for: usage-saturation lag (customers report 6-12 month adoption curves) compressing seat-expansion economics.
  5. Intelligence attach rate. Platform-expansion proof point. <20% attach = “CLM 2.0” positioning is aspiration; >40% = data gravity flywheel is real. Vantage-tier accounts (all 3 products deployed) are the upper bound; mid-market default appears to be Review-only.
  6. Series C terms and timeline. At $355M post and ~$15–25M implied ARR, Ivo needs 3x+ ARR growth or a path to $100M+ ARR before raising. A flat or down round signals bundling compression is outpacing growth. Watch for: Harvey or Legora “in-house tier” pricing below $300/user/month. Timeline: late 2027 at current burn.
Source: 4 direct Ivo-customer expert calls (May 2026); Altis category memos; public filings and press releases
11

Sources

Expert Calls

  • Direct Ivo-customer calls (May 2026, N=4 transcripts / 3 distinct voices): Chief Legal & Global Procurement Officer | Mid-Market Tech (recurring — 2 samplings Jan 18 + Jan 21, 2026, treated as N=1); Commercial Counsel | Large Consumer Internet Co.; Director / AGC | Data Center Infrastructure Co.
  • 20+ tangential expert calls across legal AI (11 GC AI, 6 Harvey, 3 Spellbook, 2 Wordsmith, 1 Sandstone, 1 Legal AI Industry)

Direct Coverage

  • 8 direct Ivo items (LegalTech Hub briefing, product demos, GOLT TALK Ep. 24)

Public Interviews & Podcasts

  • Legally Disrupted Ep. 39 (Min-Kyu Jung, CEO Ivo)
  • Diaspora.nz S2 E7 (Min-Kyu Jung)
  • Fika Pour Over (Feb 2025, Min-Kyu Jung)

Altis Category Memos

  • Altis Wordsmith memo (Apr 2026) — SAM sizing + UX-driven adoption framework
  • Altis Harvey-v-Legora memo (Apr 2026)
  • Altis GC AI memo (Apr 2026)
  • Altis Sandstone memo (Apr 2026)

Note: Altis did not have access to Ivo management team, internal documents, financials, or customer retention data. ARR reverse-engineered from public valuation.

Public References

      12

      Legal Notices

      13

      Thank you

      14