DeepIP
by DeepIP
AI patent drafting and prosecution that lives inside Microsoft Word
DeepIP is a vertical AI platform for patent work, built for IP law firms and corporate patent departments. It drafts applications, assesses patentability, searches prior art and answers office actions, and runs as a Microsoft Word add-in tuned on a firm's own prior filings so first drafts come back in the firm's house style.
DeepIP is a vertical AI platform for patent work, aimed at IP law firms and corporate patent departments rather than at general-purpose legal teams. Its distinguishing choice is workflow placement: the drafting assistant runs as a Microsoft Word add-in, so attorneys stay in the document they already file from, and the models are tuned on a firm's prior applications to reproduce its templates, tone and jurisdictional preferences instead of producing generic prose an attorney must rewrite. Beyond first-draft generation of claims, specifications and abstracts with USPTO and EPO compliance checks, the platform now spans patentability assessment with novelty scoring, semantic prior-art search across an indexed corpus the vendor puts at 150 million patents, office-action response drafting with rejection mapping, freedom-to-operate and invalidity claim-by-claim analysis, and patent drawing generation — plus Outlook and docketing-system integrations and a REST API. Security is positioned for regulated IP work: SOC 2 Type II, ISO 27001, ISO 42001, GDPR and German §203 StGB, hosted on US Azure with no customer data used for model training. Founded in 2024 by François-Xavier Leduc and Édouard d'Archimbaud, who previously built the data-labelling company Kili Technology, DeepIP raised a $15 million Series A in 2025 led by Resonance and a $25 million Series B on 2 March 2026 co-led by Korelya Capital and Serena, taking total funding to $40 million across roughly nine months. It reports 400 law firms and corporate IP teams across 25 jurisdictions, more than 40,000 matters processed and tenfold ARR growth over 18 months — all unaudited vendor figures — with named customers including Philips, Dexcom, Greenberg Traurig, Morgan Lewis, Nixon Peabody, Mewburn Ellis and Schwegman Lundberg & Woessner. In June 2026 it acquired PatentMaker, a German drafting tool used by around half of Germany's leading IP firms plus Siemens and Infineon, buying its way into Europe's largest patent market.
Heads of IP at law firms and corporate patent departments filing at volume in software, mechanical or process fields, who want drafting speed without moving attorneys out of Microsoft Word.
First drafts of applications and office-action responses come back in the firm's own style, with the vendor claiming up to two hours a day returned per attorney.
At a Glance
- Category
- Industry & Government
- Pricing
- Contact for pricing, Subscription
- Target Market
- Patent Attorneys, Heads of IP, General Counsel, Corporate IP Departments, IP Law Firm Partners
- Deployment
- Cloud-only, API-based
- Founded
- 2024
- Headquarters
- New York, United States
- Team Size
- 11-50
- Customers
- 400 law firms and corporate IP teams across 25 jurisdictions, 40,000+ matters processed (March 2026, vendor-stated). Named: Philips, Dexcom, Greenberg Traurig, Morgan Lewis, Nixon Peabody, Mewburn Ellis, Cantor Colburn, Schwegman Lundberg & Woessner.
Key Features
- ✓Word-native drafting assistant
Runs as a Microsoft Word add-in so attorneys draft and review in the same document they file from, with no tool switching.
- ✓Firm style matching
Learns from a firm's prior applications to match templates, tone and jurisdictional conventions, cutting the rewrite an attorney would otherwise do.
- ✓Prior-art and patentability analysis
Semantic search over a stated 150 million indexed patents with relevance ranking, plus novelty and non-obviousness scoring before filing.
- ✓Office-action response drafting
Parses examiner rejections, maps them to claims and drafts structured counter-arguments with case-law citations.
- ✓FTO and invalidity mapping
Maps claims against a technology claim by claim to flag infringement risk and produces exportable reports for clients.
- ✓Regulated-industry security posture
SOC 2 Type II, ISO 27001, ISO 42001, GDPR and §203 StGB on US Azure, with no client data used for training.
Capabilities
Use Cases
- •Disclosure to first draft
An attorney feeds an inventor disclosure into Word and gets claims, specification and abstract back in the firm's house style for review.
- •Office-action turnaround
The platform reads an examiner rejection, maps it against the claims and drafts a structured response the attorney edits rather than writes.
- •Pre-filing go/no-go
In-house IP counsel runs patentability and prior-art analysis before committing outside-counsel budget to an invention disclosure.
- •Freedom-to-operate review
A product team's launch is checked claim by claim against third-party patents and an exportable risk report goes to the business.
- •Cross-jurisdiction filing
A portfolio is prepared for USPTO, EPO, PCT and German filings with jurisdiction-specific conventions applied automatically to each draft.
Ideal For
Best For
- ✓Drafting patent applications, claims and specifications from invention disclosures
- ✓Responding to USPTO and EPO office actions with structured rejection analysis
- ✓Semantic prior-art search and pre-filing patentability assessment
- ✓Freedom-to-operate and invalidity claim-by-claim mapping
- ✓Corporate IP departments standardising drafting quality across outside counsel
Not Ideal For
- ✗Biotech and chemistry practices — an independent comparison rates its capability there well below its software, mechanical and process patent work
- ✗Small firms and solo practitioners: pricing is quote-only and independently estimated at roughly $350-420 per user per month, high for the category
- ✗Teams that need mature end-to-end prosecution today, since the office-action module is reported as newer and less developed than the drafting core
- ✗General corporate legal work — this is patents only, not contracts, litigation or compliance
Deployment
Market & Ratings
400 law firms and corporate IP teams across 25 jurisdictions, 40,000+ matters processed (March 2026, vendor-stated). Named: Philips, Dexcom, Greenberg Traurig, Morgan Lewis, Nixon Peabody, Mewburn Ellis, Cantor Colburn, Schwegman Lundberg & Woessner.
Market Analysis
Pros
- ✓Word-native delivery removes the adoption barrier that kills most legal-tech rollouts — attorneys never leave their drafting environment
- ✓Full patent lifecycle in one platform, extended in June 2026 by acquiring German incumbent PatentMaker
- ✓Unusually strong compliance stack for a two-year-old company: SOC 2 Type II, ISO 27001, ISO 42001, GDPR and §203 StGB, with no training on client data
- ✓Credible enterprise references including Philips, Dexcom, Greenberg Traurig and Mewburn Ellis, and $40M raised in about nine months
- ✓Broad jurisdictional coverage (USPTO, EPO, PCT and, post-acquisition, the German market) described in an independent comparison as among the widest available
Cons
- ✗Weak for biotech and chemistry patents relative to software, mechanical and process work, per an independent tool comparison
- ✗The office-action module is reported as newer and underdeveloped next to the drafting core, so full prosecution coverage is aspirational rather than proven
- ✗Priced high for the category — independently estimated at $350-420 per user per month, with no published list price to check it against
- ✗Headline metrics (tenfold ARR growth, 50-70% drafting time saved, 40,000+ matters) are vendor-stated and unaudited; the Series A coverage explicitly noted no independent validation of the time-savings claim
- ✗No usable public review footprint — G2 blocks unattended access, and Hacker News, Product Hunt and Reddit have nothing on the product, so buyer references must come from the vendor
- ✗Hallucination risk is acknowledged rather than eliminated: the company says it regularly tests models to minimise it, which still leaves attorney review as the control
Pricing
Enterprise (quote-only)
Contact for pricing
- ✓Word add-in drafting
- ✓Prior-art search and patentability assessment
- ✓Office-action responses
- ✓FTO and invalidity analysis
- ✓Outlook and docketing integrations
- ✓REST API
DeepIP publishes no list price — like Solve Intelligence, Patlytics and Clarivate's Rowan Patents, it is demo-and-quote only, so procurement cannot benchmark without entering a sales cycle. The one public estimate comes from a competitor-run but independently hosted comparison, which puts DeepIP at roughly $350-420 per user per month and calls it expensive relative to the category; that figure is an estimate, not a vendor quote, and should be treated as an order-of-magnitude anchor only. Expect seat-based annual contracting for a firm-wide rollout, with prior-art search volume and jurisdictional coverage as the likely negotiation levers.
Security & Compliance
Connect
Sources
This page was written from 7 sources, 6 on domains other than deepip.ai.
- 1.deepip.ai — deepip.aivendor
- 2.ipwatchdog.com — ai deepip raises 15m harvey patent law
- 3.juve-patent.com — deepip enters german market by acquiring patentmaker
- 4.patentbots.com — patent drafting tools comparison
- 5.finance.yahoo.com — deepip developed ai assistant drafting 080001372
- 6.tech.eu — deepip secures 25m series b to embed ai across the patent li
- 7.artificiallawyer.com — deepip raises 25m for ai driven patent work
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Ambience Healthcare
Ambient AI clinical documentation and point-of-care coding platform for health systems
Abridge
Ambient AI clinical documentation and care intelligence for health systems, deeply integrated with Epic
CADDi
AI data platform that structures fragmented manufacturing drawings, specs and purchase history
Faro Health
Agentic AI that turns clinical trial protocols from static documents into structured development intent