IP Author

IP Author

0.0 (0)
Technology
Weekly

IP Author Podcast brings practical insights for patent drafting professionals, patent prosecution experts, and in-house IP teams. Explore strategies, drafting techniques, prosecution trends, and real-

Oct 8 · Video

The USPTO's First AI Predicated Decision and What It Means for Patent Drafting | IP Author Ep. 10

The USPTO's first AI-predicated discipline order involves a patent attorney registered since 2002 who drafted a claim construction with one generative AI tool and reviewed it with another. The citations pointed to parts of the specification that did not exist. He caught the error about a day later and fixed it, the court did not sanction him, and the Office of Enrollment and Discipline reprimanded him publicly anyway. Elizabeth Shipsides of HSML Intellectual Property Law, a former USPTO examiner who has worked in-house and at firms, joins Saskia van Reineveldt and Sumer of IP Author to explain why the duty of reasonable inquiry under 37 CFR 11.18 stays with the signer, whatever tool wrote the text. Her read is that the order points to a process failure more than a dishonest act. The likeliest cause was inconsistent inputs: a reviewing tool that never received the specification has nothing to check paragraph numbers against. Sumer ties this to Article 123(2) EPC, where a general model without the application as filed cannot know it is working inside a closed universe, while a tool built for prosecution can have that constraint engineered in. Shipsides's practical rules are to keep an audit trail of AI use and paragraph checks, draft and review in the same place with the documents loaded once, and use preset prompts. They also cover why general models drift as context piles up, why tools that start each matter clean hold style and scope better, and what happens when inventors run her drafted applications through their own AI tools and return misnumbered figures and restructured text. On the division of labor, Shipsides keeps claims and drawings for herself and often rewrites or cuts what AI produces for diagram descriptions and background sections. She finds AI most useful for brainstorming office action responses, where matching its findings works as a second pair of eyes, and least reliable on strategy. Claude tends to push toward arguing when she, as a former examiner, would amend and move prosecution forward. She also covers color-coded claim charts, support checks for amendments, and why ChatGPT writes warmer client emails than Claude. Her verdict is that AI raises the quality of the work, though she is not sure it saves time. Published on Subwavehttps://subwave.app/@ipauthor/post/the-usptos-first-ai-predicated-decision-and-what-it-means-for-patent-drafting--ip-author-ep-10
Sep 29

Can AI Ship Secure Code? IP Author Podcast

Samir Raiyani and Saskia van Ryneveld sit down with Disha and Saurabh from Titan Dev to unpack a question every CTO is now asking out loud: can AI actually ship code you'd trust in production and how do you keep it secure while it does? Titan Dev is doing something specific: an AI team, not a single model. Planners, reviewers, brainstorming agents, guardrails, each with its own role, its own scope, its own accountability. The conversation moves from the design philosophy behind that architecture ("we treat AI as our team") into the guarantees that have to hold for enterprise clients to trust it: data isolation, deterministic tooling for the parts that can't be probabilistic, and where the line sits between what the model gets to author and what a human still owns. We also get honest about failure modes: where general-purpose LLMs break when you point them at real software delivery, why "AI does everything" is the wrong pitch, and what actually works. What we cover- Why Titan Dev builds AI as a team of specialised agents rather than one god model?- The security model, how client data stays isolated end to end- Deterministic vs. generative: which parts of the coding process should NEVER be probabilistic- Where general models still fall short in software engineering- From "measure twice, cut once" to shipping v1 in January, Titan Dev's build story- Guardrails, review agents, and the trust question for enterprise adoption- What this playbook means for the IP / patents world For CTOs, engineering leaders, and IP professionals thinking about where agentic AI actually earns its place in critical workflows. 00:00 Cold open — "we treat AI as our team"00:45 Host intro (Saskia) — meet Titan Dev02:30 What Titan Dev builds & who it's for05:00 The multi-agent design: planner, reviewer, brainstorming agent10:00 Why not a single god-model? Skills, rules, harnesses15:00 Data isolation — how client data stays with Titan Dev20:00 Shipping v1 in January — the build story23:00 "Measure twice, cut once" — deterministic vs. generative work26:00 Where general models still break in software delivery30:00 What the platform can actually do end-to-end35:00 The trust question: what enterprise clients demand40:00 Guardrails & review agents in practice45:00 Lessons for the IP / patents world49:00 Wrap-up — what's next for Titan Dev49:45 Sign-off #IPAuthor #AISecurity #Cybersecurity #AgenticAI #SecureCode#SoftwareDevelopment #TitanDev #LegalTech #AIforEngineering#MultiAgent #AIagents #DevSecOps Published on Subwavehttps://subwave.app/@ipauthor/post/can-ai-ship-secure-code-ip-author-podcast
Sep 29 · Video

Where General AI Still Fails at Patent Drafting | IP Author Podcast Ep. 8

Episode 8, Abhay (IP Author) sits down with Saskia van Ryneveld to unpack a question every patent professional is quietly asking: general purpose LLMs claim to be "good at everything", so why do they still fall short when you point them at a real patent specification? The conversation moves past the demo-day hype and into the mechanics that separate a passable draft from a fileable one: the narrative architecture a spec has to carry, the invariants an examiner will test, and the specific failure modes we keep seeing when a generic model is asked to do specialist work. What we cover• Why a patent isn't a document, it's a story, and every paragraph is doing narrative work• Where general models handle the surface but drop the through-line• The "good at everything" trap: benchmarks vs. the bar patent attorneys actually apply• Specialist vs. generalist AI: when the frontier model is the right tool, and when it isn't• What "AI-assisted drafting" needs to look like before firms can trust it in prosecution• Practical guidance for attorneys evaluating patent-drafting AI today For patent attorneys, agents, IP operations leads, and anyone weighing which class of AI actually belongs in the prosecution workflow. Listen & subscribe• Website: https://ipauthor.com/podcast/• Apple Podcasts: https://podcasts.apple.com/us/podcast/ip-author/id1896586525• Spotify, Amazon Music, YouTube, search "IP Author" Chapters01:00 The "good at everything" problem08:00 A patent is a story: narrative architecture in a spec18:00 Where general models break, concrete examples30:00 General vs. specialist AI for prosecution work42:00 What "AI assisted drafting" needs before firms can trust it55:00 Advice for attorneys evaluating AI today60:30 Wrap-up #IPAuthor #PatentDrafting #LegalAI #GenerativeAI #PatentLaw #PatentProsecution #AIforLawyers #Patents #IPLaw #LegalTech Published on Subwavehttps://subwave.app/@ipauthor/post/where-general-ai-still-fails-at-patent-drafting--ip-author-podcast-ep-8
Jul 17 · Video

IP AI Business Landscape: Inside the AI Patent Vendor Explosion Episode 7

Samir Rayani just returned from a US IP conference where the demo floor looked nothing like the one he walked a year ago in Europe. Nearly every booth was an AI vendor. The legacy players had retreated to bronze sponsorships, replaced by well-funded new entrants, some backed by tens of millions in capital and apparently not shy about spending it. What surprised him most was the velocity: barriers to entry are low, capital is abundant, and customers have largely stopped debating whether to adopt AI. They want it, at scale, and they do not want to build it themselves. The competitive picture now resembles a classic land grab. Vendors are broadly comparable in what they offer, customers are ready to buy, and execution speed on sales, product, and hiring will determine who surfaces on top. Samir is candid that the capital advantage he had planned to use as a differentiator has been neutralized by how much money has flooded the sector. On the talent side, skilled patent professionals are leaving law firms and joining AI startups across the board, which raises the quality of every serious player in the field and makes coasting on any early lead impossible. The bubble comparison to 1999 does not hold cleanly: every funded company in this space has some revenue, and consolidation is already happening through acquirers like Harvey and Solve Intelligence absorbing players without enough momentum to survive independently. The more consequential shift is what Samir calls the second-order effects. Claim charts that once required significant billable hours are now generated at scale by AI, flooding the patent licensing market with analysis that previously carried real cost. The economics between licensors and licensees are already moving. Further out, collapsing creation costs in adjacent fields like animated film and commercial content point toward entirely new formats that nobody has predicted yet, the same way nobody predicted TikTok or Uber from the original premise of mobile internet. The first-order disruption is priced in; what reorganizes around it is still open. Published on Subwavehttps://subwave.app/@ipauthor/post/ip-ai-business-landscape--inside-the-ai-patent-vendor-explosion-episode-7
Jul 1 · Video

Patent Examiner Insights Powered by AI | IP Author's New Data Tool Episode 6

Samir Rayani, founder and CEO of IP Author, walks through insights.ipauthor.com, a new public site built on a straightforward but consequential idea: instead of waiting for users to ask questions, AI agents process every pending U.S. Office Action continuously, build examiner profiles from full prosecution histories, and surface the results as a living dataset anyone can access. The distinction Samir draws is between raw data, human-labeled data, and a third category he calls agent-generated data. The Insights site is IP Author's first major public output in that third category. The examiner intelligence is where it gets specific. Using a live IBM Office Action as an example, the site surfaces the assigned examiner's grant rate, her lift from interview requests, her history of rejection types, and a ready-to-use strategy recommendation. In this case, a non-statutory double patenting rejection resolved by inserting a cross-reference to a related application section, with all independent claims surviving. The agents had already identified the pattern across similar rejections and generated the fix before the attorney opened the file. Portfolio managers and in-house counsel can also pull company- and law-firm-level views: which Office Actions are pending, which are easiest to overcome, and what the time and cost picture looks like. Basic access is free, no sign-up required. Deeper analytical results, including full strategy recommendations and risk assessments, are available with an email registration. For teams building internal tools, IP Author offers an MCP server that gives agents from Claude or other platforms direct access to the full dataset. The site is live now at insights.ipauthor.com, and feedback goes to support@ipauthor.com. Published on Subwavehttps://subwave.app/@ipauthor/post/patent-examiner-insights-powered-by-ai--ip-authors-new-data-tool-
Jun 17 · Video

Who’s Actually Using Legal AI? Early Adopters, Skeptics & the Future of Patent Law Episode 5

Saskia van Ryneveld and Sumair, VP of Business Development at IP Author, dig into a distinction that cuts to the heart of how patent professionals are actually engaging with AI: the difference between being an AI tourist and an AI resident. Tourists experiment, take a few tools for a spin, and return to their old workflows. Residents integrate AI deeply enough that it changes how they practice, not just how fast they complete individual tasks. The conversation covers how patent attorneys are using AI as a skill multiplier rather than a substitute for legal expertise, and why the distinction matters for the quality of output. Sumair also walks through the differences he observes between European and US patent professionals in their pace and style of adoption, including where resistance tends to come from and what drives genuine uptake. The professionals who are shaping how legal AI tools actually get built and refined are the ones using them most critically today. That feedback loop between power users and developers is where the real evolution of AI in patent practice is happening. Published on Subwavehttps://subwave.app/@ipauthor/post/whos-actually-using-legal-ai-early-adopters-skeptics--the-future-of-patent-law-episode-5
Jun 1 · Video

Why Open Source AI Changes Everything for Patent Attorneys Episode 4

How is open source AI changing the future of patent law and legal technology? 🔗 Learn more about IP Author: https://ipauthor.com In this episode of the IP Author Podcast, Saskia van Ryneveld, Samir Raiyani and Sumair Riyaz explore:,Open source legal AI,The future of patent workflows,AI-generated innovation,Why “the blank page” may disappear,The evolving role of patent attorneys,AI as a creative and technical collaborator,How platforms like IP Author are building AI workflows for IP professionals. The conversation also explores a fascinating question:If AI can generate endless possibilities, what becomes the uniquely human contribution? Published on Subwavehttps://subwave.app/@ipauthor/post/why-open-source-ai-changes-everything-for-patent-attorneys-1
May 28 · Video

The Rise of AI Agents: What Law Firms Need to Know. IP Author Podcast Episode 3

Samir Rajani presented AI agents to a large law firm and found 70-80% of the room apprehensive or hostile. His advice: stop waiting, find where agents multiply your output, and keep your judgment at the center. Published on Subwavehttps://subwave.app/@ipauthor/post/the-rise-of-ai-agents--what-law-firms-need-to-know-ip-author-podcast-episode-3
May 22 · Video

How AI Agents Are Transforming Patents & Legal Work: IP Author Podcast Episode 2

Samir Raiyani, founder of IP Author, breaks down why Claude Code is generating serious attention across the tech industry and what it means for knowledge workers building with AI today. Claude Code operates through autonomous "agents" that can collaborate, self-correct, and complete multi-step tasks without constant human input. Legal and IP professionals are adopting these tools faster than most industries expected, and Samir explains the mechanics behind why that adoption is accelerating. Patent analysis, office action review, and competitor portfolio forecasting are no longer tasks that require weeks of manual research. AI agents can now handle structured legal reasoning at a speed and scale that changes how IP strategy gets built inside companies. Samir walks through what the Agentic IP Author is already capable of today, including live demonstrations of the workflows it can run end to end. The broader question underneath all of this is what "autonomous AI" actually means in practice for legal teams and businesses. Predicting a competitor's patent filings before they happen, surfacing claim gaps in existing portfolios, and drafting responses to office actions are concrete tasks these tools are already performing. Samir covers where the technology sits right now and what the next capability thresholds look like. Published on Subwavehttps://subwave.app/@ipauthor/post/how-ai-agents-are-transforming-patents--legal-work--ip-author-podcast-episode-2
May 12 · Video

Agentic AI for Patents: The Future of Patent Drafting, Prosecution & IP Strategy

Claude Code and multi-agent AI systems are changing the mechanics of patent drafting in ways that go beyond simple automation. Tools that can run agentic workflows now handle tasks like prior art search, claim charting, and whitespace analysis, reducing the manual overhead that has historically slowed down patent professionals at every stage of prosecution. Office action responses are one of the clearest areas where AI is delivering measurable value. Agent-based systems can analyze examiner rejections, surface relevant prior art, and generate draft arguments, giving patent attorneys and agents a faster, more structured starting point. The same underlying capabilities apply to portfolio development, where AI can scan a technology space and flag gaps worth protecting. The episode also covers the security and risk considerations that come with deploying advanced AI models on sensitive IP work, including how to think about data exposure and model reliability when the stakes involve client confidentiality and claim validity. These are practical concerns for any IP professional evaluating whether and how to bring these tools into their practice. Published on Subwavehttps://subwave.app/@ipauthor/post/agentic-ai-for-patents--the-future-of-patent-drafting-prosecution--ip-strategy

10 Episodes

IP Author Podcast brings practical insights for patent drafting professionals, patent prosecution experts, and in-house IP teams. Explore strategies, drafting techniques, prosecution trends, and real-

Creator

IP Author
Years Active

2026
Episodes

10
Rating

Clean
Show Website

IP Author