[TheFounder] Redefining IP Strategy with Simon Lee, Co-Founder of Patenty.ai

Sangmin Simon Lee is the co-founder of Korea-based Patenty.ai, a platform building AI-powered patent drafting solution to help patent professionals make faster, more sophisticated decisions. Patenty.ai distinguishes itself by positioning AI not as an automation tool that replaces expertise, but as a medium to amplify human judgment. The core vision is to elevate speed, accuracy, and strategic insight simultaneously by refining the complex processes of patent research and strategy formulation.

Simon’s career is defined by continuous venture building. As the founder and CEO of WeeklyVentures, a serial venture studio, he has built and validated over 20 products, leading to successful outcomes such as the spin-off and sale of SoulCanvas and the commercial launch of Patenty. He has further explored the startup ecosystem and AI applications through experimental projects like Personia, AMPulse, CADlingo, and Weeklylab.

His experience as Regional Director Asia for the German VC AM Ventures where he reviewed advanced manufacturing startups and his collaboration with global manufacturing firms at EOS GmbH, provided him with a multi-dimensional perspective on the practical value of technology, industry, and IP. This background is exactly why Patenty.ai was designed as an intelligent decision-making platform that works in the field, rather than just a simple patent specification generator.

In this interview, we explore how Simon Lee’s multi-faceted career converged into the mission of Patenty.ai and how he is redefining the role of patents and intellectual property in the age of AI.

Q. Welcome to The Founder interview. You have an extensive background as a serial venture builder and an investor. Given your deep expertise in both tech and the startup ecosystem, how would you define Patenty.ai ?

Thank you for the warm welcome. Patenty.ai is a professional AI workspace that liberates patent attorneys from repetitive, manual drafting labor and returns them to their core mission: claim strategy and IP portfolio design.

Q. The patent and IP sector is notoriously difficult for AI application. How does it differ structurally from other knowledge work, and why did you choose this field?

While AI advancements have lowered the barriers to entrepreneurship, general-purpose services are becoming harder to sustain. We saw an opportunity in vertical domains where expert feedback is essential, external data is difficult to collect, and a deep understanding of on-site workflows is required.

Patent work is complex knowledge labor that must simultaneously solve three challenges: technical understanding, legal interpretation, and natural language processing. You have to transform the abstract concept of an "invention" into the logical structure of a "legal right." We believed general LLMs alone could not guarantee the logical integrity and timeliness required in this field-and paradoxically, that is where the opportunity for innovation lies.

Q. Patenty.ai defines itself as "Assistive AI" rather than "Replacement AI." What is the distinction?

The output of AI usually sits somewhere between an expert and a layperson. A user without specialized knowledge might be amazed by an AI's answer, but an expert will see many flaws.

We adopted a collaboration model similar to that between a junior and a senior patent attorney. The AI acts like a junior, processing vast data and creating initial drafts, while the senior attorney verifies the integrity and makes strategic judgments. By maintaining a Human-in-the-Loop structure, we help attorneys focus on high-value work.

Q. How do Patenty.ai's core features map to the patent workflow?

Unlike many AI tools that are fragmented, we manage patent work as a single, connected context:

1. Concept Decomposition & Drafting: We analyze inventions into structured "concept nodes" and draft claims using a block-based system. Users can then generate full specifications and guide the refinement through direct edits or AI instructions.

2. Prior Art Search: Integrated with IP5 patent data, our semantic search quickly identifies similar inventions. Users can selectively save documents or generate prior art reports.

3. Official Action (OA) Response: When an examiner issues a rejection, the AI analyzes the logic based on the technical context and suggests defense strategies and options.

Q. Why do you emphasize domain-specific AI over general-purpose LLMs?

General LLMs generate plausible sentences based on probability; they struggle with the logical rigor required by patent law.

The core of Patenty.ai is our Concept-Native Engine and IP5 Data-based RAG (Retrieval-Augmented Generation). We refine an invention into structured knowledge so the AI understands the technical architecture before generating text. Furthermore, we only reference verified IP5 data and legal documents, and we can even replicate a specific organization's writing style and tone - a level of personalization that general models cannot match.

Q. In terms of market strategy, why did you focus on Asia from the start?

The Asian region (China, Korea, Japan) accounts for roughly 60% of the world's patents. Despite being the center of global IP filings, it remains massively underserved by Western solutions due to complex language barriers and unique practices that differ from the West. Moreover, the workload per attorney in Asia is among the highest in the world, making the need for productivity innovation through AI most urgent.

Q. What is the biggest differentiator between Patenty.ai and its competitors?

There are two clear differences: Form Factor and Coverage.

• Workspace, not Plugin: Some competitors operate as MS Word plugins. While familiar, they are limited by the constraints of a text editor. Patents are not just text; they are structures where claims, drawings, and prior art are organically linked. Patenty.ai provides a web-based workspace - unified patent drafting environment - where you can visualize concepts in blocks and compare data in real-time.

• Global IP5 Coverage: Most existing solutions are designed for Western patent environments. However, Asian patent laws have vastly different drafting practices and examination standards. We are building native logic that reflects the data and workflows of Korea, China, and Japan to provide a truly global IP5 capability.

Q. How do you see patent work changing in the next 3 to 5 years?

The cost of drafting specifications will eventually approach near-zero. At that point, IP competitiveness will be determined not by who writes faster, but by who uses AI to build the best portfolio strategy and preempts technical white spaces.

Q. As we wrap up, we’d like to shift our focus to the bigger picture. With your unique perspective on how technology reshapes industries, is there one truth about the future of AI that many people are overlooking?

People worry about AI taking jobs, but the more fundamental question is: when work disappears, what do we define ourselves by? Patent attorneys face this too. Once drafting is automated, their identity must shift from "the person who writes well" to "the person who judges well." The gap will widen between organizations that embrace this transition and those that resist it. Patenty.ai exists precisely at that inflection point - not to replace the attorney, but to accelerate this redefinition.