AI-Driven Intellectual Property Management Advances in Silicon Valley (Amplified, Triangle IP, PatSnap, Harvey, Anaqua)

The Three-Layer Structure of IP AI — Why VCs Are Paying Attention Now

To understand the IP AI market, it is necessary to organize three layers that follow the patent lifecycle.

Traditionally, patent work was a linear process: "write an idea on paper → hand it to a lawyer → spend several months filing → spend several years in examination." With AI intervention, each stage of this process is being dramatically compressed and elevated. The reason Silicon Valley VCs are paying attention to this space is clear. The thesis that a16z articulated — "for every $1 spent on software, $6 is spent on services; AI will eat that $6" — applies precisely to the world of intellectual property. The cost of filing a single patent averages $10,000–$30,000 (approximately ¥1.5M–¥4.5M) in the United States, and it is not uncommon to exceed $100,000 (approximately ¥15M) when international filings are included. The majority of that cost is tied to service fees linked to attorneys' hourly rates, which means the room for AI-driven efficiency gains is enormous.

---

Layer 1: Creation — AI instantly evaluates engineers' ideas

Triangle IP: "TIP Tool" — Democratizing Invention Disclosure

Company Overview

Product Core

Triangle IP's TIP Tool™ is a cloud platform that manages the entire lifecycle of an invention — from the moment an idea is conceived through to patent grant. Its defining feature is dramatically lowering the barrier to invention disclosure.

Traditionally, when an engineer came up with a new technical idea, conveying it to the internal IP department required filling out a complex Invention Disclosure Form. Many engineers found this process burdensome, and valuable ideas quietly disappeared inside the organization — the so-called "dark matter of ideas" problem.

TIP Tool provides an intuitive drag-and-drop interface that lets engineers submit ideas across departments. Once submitted, a machine learning model predicts the Art Unit (examination division) assignment based on historical USPTO examination data and automatically generates an early-stage patentability outlook. In other words, when an engineer asks "Does this have a shot at becoming a patent?", the AI returns an initial answer in seconds.

Cases have been reported where IP departments achieved a 30% improvement in patent allowance rates. This results from AI-powered early screening that avoids wasting resources on low-patentability ideas, enabling concentrated investment in ideas with stronger patentability.

Data is encrypted on AWS servers, and security is substantially improved compared to email-based disclosure.

VC Perspective

Triangle IP is currently a bootstrapped company that has not raised significant venture capital. However, this positioning carries interesting implications for VCs. An invention disclosure platform sits at the "front door" — closest to the enterprise R&D function. A company that controls this layer can build an overwhelming advantage through data integration with downstream filing and management processes. With major platforms such as PatSnap and Anaqua looking to strengthen this layer, a specialist player like Triangle IP is a credible M&A target.

---

Amplified: An AI That Has Read 140 Million Patents — and Understands "Meaning"

Company Overview

Product Core

Amplified's uniqueness lies in its proprietary language model trained on more than 140 million patent documents. Co-founder Chris Grainger conducted PhD research applying neural networks to patent data within the field of innovation economics, and that work forms the foundation of Amplified's technology.

Conventional keyword or Boolean searches only reach the "surface" of patent documents. Searching for "image recognition," for example, will miss prior art described using synonyms such as "visual pattern analysis" or "optical feature extraction." Amplified's language model understands not just what a patent says, but what it means — enabling semantic search based on conceptual similarity.

Key Features:

• Dual-mode search: Toggle between classic semantic search and conceptual-similarity neural search
• Interactive map: Clusters and visually maps search results by conceptual similarity, automatically grouping related patents
• Collaboration: Share projects across teams, annotate patents, and discuss via chat
• Guided workflows: Supports core tasks including invention disclosure, invalidity searches, and monitoring
• Multilingual support: Full-text coverage from 45 patent offices across 100+ jurisdictions, including non-Latin script search

VC Perspective

Amplified is a seed-stage company whose investors include Tachi.ai Ventures, an early-stage VC bridging Japan and Europe. Adoption by 50+ large enterprises is noteworthy evidence of product-market fit. Competition in the patent search market is intensifying — IPRally (Finland, graph AI), Patlytics (cumulative funding ~$65M), and DeepIP (cumulative funding ~$40M) are among the growing field — but Amplified's "proprietary model trained on 140 million patents" represents a genuine differentiator. As it looks ahead to a Series A, expanding data partnerships — such as its collaboration with IFI Claims Patent Services — will be a critical factor.

---

Layer 2: Practice — Generative AI Automatically Creates Patent Documents

Harvey: The $11 Billion Legal AI Giant Makes a Full Push into IP

Company Overview

Full-Scale Expansion into IP Practice

Harvey initially grew rapidly as a "legal AI operating system" covering general legal work such as M&A, compliance, and contract review. However, entering 2026, it has accelerated its full-scale push into patent and intellectual property.

In February 2026, Harvey released five new workflow templates for IP and patent litigation.

1. Automated Infringement Claim Chart Generation

Work that previously required associates to spend hours manually mapping patent claims against technical documents for accused products is now automated by AI. The system extracts product features corresponding to each claim element, cites specific page numbers, and flags uncertain elements for human review. Output is delivered in a structured table format showing claim elements, product features, document references, and analysis notes at a glance.

2. USPTO Office Action Analysis

Analyzes rejection notices from the USPTO (§101, §102, §103, §112), structurally extracting the type of rejection, cited prior art references, and the examiner's reasoning. It also auto-generates claim amendment proposals and strategies for arguing distinctions over prior art. IP boutique firm Estrella has noted that Harvey's Vault feature — which enables reuse of successful argument structures — allows them to "produce more consistent, higher-quality responses in less time."

3. Invalidity Contention Generation from Prior Art

Takes patent documents and prior art references as input and automatically performs limitation-by-limitation comparisons. Results are output in structured tables for each claim, enabling systematic verification of coverage across multiple patents.

4. Patent License Agreement Drafting

Generates complete agreements from business terms (exclusivity type, territorial scope, royalty structure) and firm precedents as inputs.

5. Application Document Preparation

Automatically populates application templates with inventor information, assignee data, priority claims, and more.

Full Patent Lifecycle Coverage

Harvey's IP capabilities extend beyond these five workflows. The official blog post "Harvey in Practice: How IP Teams Manage the Patent Lifecycle" introduces a comprehensive IP workflow covering eight stages:

1. Invention Identification: Analyzing invention disclosures, extracting technical concepts, identifying novel aspects

2. Pre-Filing Analysis: Prior art mapping of draft claims, technology landscape analysis to identify competitive positioning and white space

3. Application Preparation: Structured drafting of application data sheets, declarations, and powers of attorney

4. Prosecution: Office action responses, distinguishing analysis over prior art

5. Post-Grant Proceedings: Brief preparation for PGR (Post-Grant Review) and IPR (Inter Partes Review)

6. Portfolio Management: Cross-analysis of portfolios by technology area, claim scope, filing date, and business unit

7. Licensing & Transactions: License agreement review and drafting, due diligence

8. FTO & Infringement Analysis: Identifying assertion candidates, drafting claim charts

The VC Perspective — The Race to Become "The Harvey of Patents"

Harvey's entry into IP has attracted significant attention in the VC world. Sequoia has called it "the most compelling example of deploying LLMs into real work," and top-tier VCs including a16z, Kleiner Perkins, and Coatue have all invested.

At the same time, VCs are closely watching the surge of startups competing to become "the Harvey of patents."

• DeepIP (Paris-based): Raised $25M in a Series B in March 2026, bringing total funding to $40M. Co-led by Korelya Capital and Serena. Adopted by 400+ law firms and in-house teams, with ARR growing 10x in 18 months. A copilot-style tool integrated as a sidebar within MS Word, supporting claim rewriting, boilerplate generation, and section summarization.
• Patlytics (Silicon Valley-based): Raised $40M in a Series B in April 2026, led by SignalFire. Total funding approximately $65M. An end-to-end AI platform covering the full patent lifecycle from invention harvesting to litigation support.
• Solve Intelligence: Raised $40M in a Series B in December 2025. Features jurisdiction-specific models for USPTO, EPO, and others, with 150+ IP firms as users.
• Ankar (founded by Palantir alumni): Building an AI platform analyzing 150 million patent applications and scientific papers, with a $20M Series A.

Harvey's strength lies in the breadth of its platform — it can layer IP on top of a foundation that already covers general legal work. However, IP is a highly specialized domain, and focused players like DeepIP and Patlytics may outperform it in depth. The "platform vs. vertical" debate is heating up among VCs, and this dynamic is reminiscent of the classic Salesforce vs. vertical SaaS battle.

---

Layer 3: Strategy — AI Advises on Competitive Analysis and Cost Optimization

PatSnap: The Full Picture of Innovation Intelligence Drawn by a Unicorn

Company Overview

The Core Product — From "Patent Search" to "Innovation Brain"

PatSnap launched in 2007 as a "patent directory," but has since evolved significantly into an AI-driven innovation intelligence platform.

The symbol of this transformation is "Hiro", an AI assistant built on a proprietary LLM.

Hiro's Key Features:

• Automated Patent Classification: AI learns the tagging logic of selected patent groups and automatically scales it to larger datasets, dramatically improving portfolio review efficiency
• AI Disclosure Document Drafting: Develops, drafts, and refines technical concepts to generate AI-drafted disclosure documents, covering Layer 1 (creation) functionality
• Patent Prosecution Navigation: Analyzes how others have overcome patent rejections through claim amendments and arguments, strengthening your own prosecution strategy
• Bridging IP and R&D: AI summaries break down the language barrier between IP and R&D teams, extracting technical insights from patents and accelerating decision-making

Strategic Analysis Features:

PatSnap's true value lies in competitive tech mapping. It visualizes competitor portfolios through citation networks, timelines, and technology domain heatmaps to support strategic decisions such as:

• Which technology areas are covered by your own R&D investments
• Which areas competitors are concentrating their filings in
• Where unexplored "white spaces" exist
• Which companies are candidates for licensing or co-development

PatSnap's platform is designed for cross-functional use across R&D, legal, procurement, and corporate strategy departments, positioning itself as an "innovation decision-making foundation" that goes beyond a simple "patent search tool."

VC Perspective

PatSnap's investor roster presents a unique composition where Asian and Western global capital intersect: SoftBank Vision Fund II, Tencent, Sequoia China (HongShan), Shunwei Capital led by Xiaomi founder Lei Jun, and Vertex Growth — reflecting the fact that intellectual property is a global competitive foundation.

Surpassing $100 million ARR is an outstanding figure for an IP AI startup. However, it is worth noting that no major new funding round has been reported since the 2021 Series E. The market is debating "when PatSnap's next round or IPO will come." If AI-driven enhancements accelerate ARR growth, an IPO or major round in late 2026 to 2027 is possible.

---

Anaqua: The Enterprise IP Management Giant Backed by Nordic PE

Company Overview

The Core Product — Injecting AI into 20 Years of Accumulated Expertise

Anaqua is an enterprise platform that provides end-to-end lifecycle management for IP assets including patents, trademarks, and designs. Fortune 500 companies and global law firms use it to manage patent portfolios ranging from tens of thousands to hundreds of thousands of assets.

AQX® 11, released in June 2024, was positioned as "the most significant platform release in 20 years," featuring full integration of AI capabilities.

AI Feature Details:

1. AI Patent Auto-Classifier™

Uses a hosted LLM to automatically map internal and external patents to a company's proprietary classification framework. Rather than simply assigning IPC (International Patent Classification) codes, the key differentiator for enterprise clients is classification aligned to each company's uniquely designed taxonomy. This enables immediate strategic analysis by business unit or technology domain, even across portfolios of tens of thousands of patents.

2. Document Auto-Processing (AI Docketing)

Incorporates Microsoft Azure AI Document Intelligence to automatically process correspondence from major patent offices worldwide. Currently supports 850+ forms, including 400 USPTO forms and 450 EPO forms. It significantly automates the docketing (deadline management) work that previously required manual data extraction and verification by humans.

3. AI Patent Summaries™

Uses generative AI to automatically produce consistent, high-quality summaries of published patents, serving as a bridge for technical staff to quickly understand patent content.

4. Agentic Workflows (Rolling out in 2026)

What Anaqua positions as "defining the future of IP management" is agentic AI. It applies LLM reasoning, planning, and autonomous action capabilities to IP workflows, with agents autonomously triggering workflows, monitoring dockets, classifying documents, and updating systems in real time.

Strategic Significance of the RightHub Acquisition

The May 2025 acquisition of RightHub added a new dimension to Anaqua's strategy. RightHub is an IP management platform designed AI-native from the ground up, targeting mid-sized law firms and corporations. By combining Anaqua's enterprise foundation (AQX) with RightHub's AI-first foundation, the company aims to build "the industry's first AI-native IP operating system."

Nordic Capital's Managing Director Aditya Desaraju described RightHub as "a perfect addition to Anaqua," signaling an aggressive investment stance in IP management for the AI era.

VC Perspective — The PE × SaaS Growth Model

A notable characteristic of Anaqua is that it is owned by Nordic Capital, a PE (private equity) firm rather than a VC. Nordic Capital completed the acquisition in February 2025 and executed the RightHub acquisition just three months later. This pace suggests that PE is moving quickly to consolidate the AI-leveraged IP SaaS market.

The case of Microsoft selecting Anaqua to "future-proof" its own IP management is a testament to credibility in the enterprise market. Detailed financial metrics are not public since Anaqua is not a listed company, but its global customer base of leading corporations and law firms, combined with Nordic Capital's backing and aggressive M&A strategy, suggests the possibility of an IPO or secondary sale within the next few years.

---

Competitive Landscape — Notable Players Beyond the Top 5

The IP AI market is expanding rapidly, and there are notable players beyond the five companies worth watching.

---

Market Size and Growth Forecast

Growth forecasts for the patent analytics market vary across research firms, but all point to strong expansion.

In Japanese yen, this represents a market of approximately ¥213 billion as of 2026, growing to roughly ¥558 billion to ¥986 billion by 2034.

Three drivers behind market growth:

1. Surge in patent filings: AI-related patent applications at the USPTO reached 45,600 per year (2023) and continue to grow. Global patent filings are also hitting all-time highs.

2. Maturation of AI technology: Advances in generative AI and LLMs have dramatically improved automated specification drafting, prior art search accuracy, and the depth of portfolio analysis.

3. Sophistication of corporate IP strategy: 68% of organizations have integrated patent analytics into their innovation strategy, shifting the role of IP departments from "cost center" to "core of strategic decision-making."

---

How Silicon Valley VCs Are Reacting

a16z: An Extension of the "Convert Labor into Software" Thesis

a16z has advanced the thesis that "for every dollar spent on software, six dollars are spent on services," and has been actively investing across vertical AI. Its investment in Harvey is emblematic of this. In a16z's 2026 "Big Ideas," a future is envisioned where AI agents autonomously execute specialized service tasks, and IP work is positioned as a quintessential use case for exactly that. For a16z, which has assembled the largest AI-focused fund in history at $20 billion, IP AI is one of the central pillars of its investment thesis.

Sequoia: An Indirect Bet on IP AI Through Its Harvey Investment

Sequoia co-led Harvey's latest round ($200M) alongside GIC, supporting the company's expansion into the IP space. Sequoia has set a goal of having more than 10 portfolio companies exceed $100M ARR by 2027, and Harvey's $190M ARR already surpasses that threshold by a wide margin. Additionally, Sequoia China (HongShan) participated in PatSnap's Series E, securing exposure to IP AI through multiple channels.

Kleiner Perkins: The Reach of Its $3.5 Billion AI-Focused Fund

Kleiner Perkins launched a $3.5 billion (approximately ¥525 billion) fund dedicated to AI startups in 2026. The firm is also an investor in Harvey and has demonstrated strong interest in vertical AI broadly, including IP AI.

SoftBank Vision Fund: A Bet on Asia-Born Innovation Intelligence

SoftBank Vision Fund II led PatSnap's Series E. Under Masayoshi Son's thesis that "AI is the greatest revolution in human history," investing in innovation intelligence—including patent data—is a logical consequence.

Nordic Capital: A PE × SaaS Integration Strategy

Nordic Capital's acquisition of Anaqua, followed immediately by the add-on acquisition of RightHub, represents a new model in which PE leads the consolidation of the IP AI market. While the approach differs from VC, the net effect is an acceleration of market consolidation.

---

Coverage by each media analyst

TechCrunch / CNBC / Bloomberg

Harvey's $11 billion valuation round was covered by all major outlets including TechCrunch, CNBC, and Bloomberg. CNBC ran a piece titled "VCs Are Spreading Their Bets Beyond Model Companies," highlighting a shift in investment toward the application layer. Bloomberg emphasized "the rapid maturation of the legal AI market."

IPWatchdog

IPWatchdog, a publication specializing in intellectual property, described DeepIP's $40 million raise as "establishing the standard for the AI patent platform market." The site published an analysis piece in April 2026 titled "AI's Inflection Point in the Patent Monetization Market," arguing that AI is fundamentally transforming how patents are valued and monetized.

Artificial Lawyer

Artificial Lawyer, a specialist site covering legal AI, reported on Anaqua's acquisition of RightHub as "the dawn of the AI-native platform era."

AlleyWatch

New York startup media outlet AlleyWatch featured Patlytics' $40 million raise, analyzing that "AI is driving a simultaneous surge in patent filings and IP litigation."

Japanese Media

Business Insider Japan reported on NEC achieving up to 94% efficiency gains in IP AI development, cutting patent research time from 22 hours to 3 hours. NEC began selling its SaaS tools and consulting services externally in April 2026, targeting ¥3 billion in revenue by the end of fiscal year 2030.

---

Japan's IP AI Trends

Japan is also accelerating the adoption of AI in intellectual property.

Policy front:

• The Cabinet Office Intellectual Property Strategy Promotion Secretariat is promoting the development of an IP ecosystem for the AI era as part of its "Intellectual Property Promotion Plan 2025"
• The Japan Patent Office (JPO) has formulated an action plan for leveraging AI technology and conducts regular surveys on the status of AI-related patent applications
• The "Guide for Rights Holders" (formulated in November 2024) outlines the expected initiatives of rights holders within the generative AI and intellectual property ecosystem

Corporate initiatives:

• NEC: Announced its "IP DX Business" in January 2026, reducing patent searches from approximately 22 hours to approximately 3 hours (up to 94% efficiency gain). External sales launched in April 2026, with a revenue target of 3 billion yen by the end of fiscal year 2030
• renue: A 2026 edition guide titled "What is Patent AI?" provides a comprehensive overview of how AI patent searches work and the latest trends in IP DX
• IP Information Fair & Conference: Implementation-stage initiatives are increasing, including exhibitions by companies that have obtained a total of 9 patents related to the utilization of generative AI

A distinctive feature of Japan is that, unlike the startup-driven model in the United States, the approach is centered on in-house development by large corporations combined with government-led initiatives. However, connections with Silicon Valley are expanding — PatSnap has established an office in Japan, and Tachi.ai Ventures, an investor in Amplified, is a VC firm bridging Japan and Europe.

---

Future Outlook — When and What Will Happen

H2 2026 (Q3–Q4)

• Harvey: Significant expansion of IP workflows is anticipated. A broadening of jurisdictions beyond the current 5 templates is expected, including PCT international filings, EPO prosecution, and WIPO coverage. Among the 25,000+ custom agents, the proportion of IP-specialized agents may surge rapidly.
• PatSnap: No major funding round has been reported since the Series E in 2021. If ARR growth has continued at ~25% YoY, ARR could reach approximately $150M in 2026, making conditions ripe for an IPO or a new round.
• Anaqua: M&A activity under Nordic Capital is expected to accelerate. Beyond RightHub, acquisitions of tools focused on specific verticals (pharmaceutical IP, standard-essential patents, etc.) are anticipated.

2027

• Market Consolidation: Consolidation through M&A or partnerships is expected among the currently fragmented IP AI startups (DeepIP, Patlytics, Solve Intelligence, IPRally, Ankar, etc.). Platform players like Harvey and PatSnap may accelerate efforts to absorb specialist players.
• Agentic AI Goes Mainstream: The agentic workflows championed by Anaqua will approach industry-standard status. A hybrid model in which humans "supervise" and AI agents "execute" becomes the norm.
• USPTO/EPO AI Adoption: Patent offices themselves will increase AI utilization to reduce the USPTO's average examination period of 23.4 months. Clarification of guidelines around AI-generated patent specifications is also anticipated.

2028 and Beyond

• Commoditization and Differentiation in IP AI: AI-assisted basic patent search and drafting will commoditize, shifting the focus of differentiation to "strategic insights" (Layer 3). The relative value of portfolio management and strategic analysis platforms like PatSnap and Anaqua will rise accordingly.
• Three-Layer Integration: An "IP AI OS" will emerge in which creation → prosecution → strategy are seamlessly connected on a single platform, consolidating today's fragmented toolchains.
• Japan Market Comes of Age: Starting from NEC's commercialization of its IP DX business for external clients (April 2026), a Japan-native IP AI ecosystem may take shape. If the FY2030 revenue target of ¥3 billion is achieved, more major companies are likely to follow suit.

---

Summary——Intellectual Property AI is moving toward "3-layer integration"

The competition in IP AI is no longer about the superiority of individual features — it is shifting toward an architectural battle over how to integrate three layers.

• Triangle IP and Amplified capture every engineer's idea at the "creation layer" and evaluate them instantly with AI
• Harvey operates at the "practice layer," where generative AI automatically drafts patent documents and frees attorneys' time for strategic work
• PatSnap and Anaqua support competitive analysis, portfolio optimization, and cost management with AI at the "strategy layer"

Silicon Valley VCs are investing in each of these three layers while anticipating that an "IP AI OS" spanning across all layers will ultimately emerge. Will Harvey push down from general law? Will PatSnap push up from innovation intelligence? Or will Anaqua leverage PE capital to drive consolidation? — This three-way competition will define the IP tech market over the next two to three years.

A future in which the cost per patent — tens of thousands of dollars — is dramatically compressed is no longer science fiction. The question is not "when" but "who will dominate the market." And the ones who hold that answer are the VCs lined up along Sand Hill Road in Silicon Valley.

---

Summary — Intellectual Property AI is moving toward "3-layer integration"

The competition in IP AI is no longer about the superiority of individual features — it is shifting toward an architectural battle of how to integrate three layers.

• Triangle IP and Amplified capture every engineer's idea at the "creation layer" and evaluate it instantly with AI
• Harvey operates at the "practice layer," where generative AI automatically drafts patent documents, freeing attorneys' time for strategic work
• PatSnap and Anaqua support competitive analysis, portfolio optimization, and cost management with AI at the "strategy layer"

Silicon Valley VCs are investing in each of these three layers while anticipating that an "IP AI OS" that cuts across all layers will ultimately emerge. Will Harvey push down from general law? Will PatSnap push up from innovation intelligence? Or will Anaqua leverage PE capital to drive consolidation? This three-way race will determine the IP tech market over the next two to three years.

A future where the cost per patent — currently tens of thousands of dollars — is dramatically compressed is no longer science fiction. The question is not "when" but "who will dominate the market." And the ones who hold that answer are the VCs lined up along Sand Hill Road in Silicon Valley.

---

Sources

• Harvey - $11B Valuation (CNBC)
• Harvey - $11B Valuation Blog
• Harvey - IP Teams Patent Lifecycle
• Harvey - 5 IP Litigation Workflows
• PatSnap Series E $300M (TechCrunch)
• PatSnap ARR $100M
• PatSnap Hiro AI Assistant
• Anaqua AQX 11 Launch
• Anaqua RightHub Acquisition
• Nordic Capital Acquires Anaqua
• Anaqua Agentic Workflows
• Anaqua - AI Reshaping IP
• Amplified.ai
• Amplified - LOT Network Profile
• Amplified - IFI CLAIMS Interview
• Triangle IP
• Triangle IP - Patent Allowance 30% Increase
• DeepIP $40M Funding (IPWatchdog)
• Patlytics $40M Series B (AlleyWatch)
• Patent Analytics Market (SkyQuest)
• Patent Analytics Market (Fortune Business Insights)
• AI Patent Tools 2026 (Patentext)
• USPTO AI Patent Trends
• NEC IP DX (Business Insider Japan)
• Patent Strategy 2026 Trends (IPEG)
• IPWatchdog - Patent Monetization AI Inflection Point
• Kleiner Perkins $3.5B AI Fund