Introducing SLW Labs

Introducing SLW Labs: Why We Built the Tools, and What Comes Next

After years of evaluating third-party AI drafting tools in production, what we observed is that the most distinctive value comes not from third-party tools, but rather from building our own tooling based on how our patent attorneys actually practice, the firm’s prosecution expertise, quality standards, and docket structure encoded in the systems themselves. SLW Labs is how we build that.

Patent prosecution is changing. Generative AI has moved from speculative to operational in two years, and most law firms now have at least one drafting or analytics tool deployed in production. SLW was among the earlier adopters, we deployed third-party tools about a year ago, and that experience taught us a lot about the limitations of those third-party tools. The most important thing it taught us is that the next decade of competitive advantage in patent prosecution will not come from buying the right tool. It will come from building the right tools.

That is the thesis behind SLW Labs.

What SLW Labs Is

SLW Labs is the firm’s internal innovation platform. It is a hub-and-spoke program. A small core team maintains shared infrastructure, authentication, audit logging, matter isolation, retrieval, and the standards every Labs product is held to. Distributed builders across the firm, attorneys, technologists, and an external development partner working under SLW direction, create domain-specific workflows on top of that platform. The model is designed so that innovation happens across the firm rather than being owned by a central team, while quality, security, and governance hold uniformly.

This matters because the alternative, buying commercial tools and adapting workflows around them, caps how much advantage any one firm can build. Commercial tools are designed for the average law firm’s average workflow. Patent prosecution at SLW’s scale, across the firm’s docket structure, prosecution patterns, and quality standards, is not average. The tools we build encode what we know.

The hub-and-spoke model rests on a single piece of infrastructure: SLW Sandbox. Sandbox is what the small core team maintains, the shared authentication, audit logging, matter isolation, retrieval, security primitives, integration libraries (USPTO and EPO feeds, scientific OCR pipelines, the firm’s docket structure), and reusable components that every Labs product depends on. Spokes work in Sandbox to build domain-specific workflows. Without Sandbox, every spoke would have to rebuild the foundation; with it, attorneys focus on the workflow that only patent attorneys can design.

Sandbox is also a structured program for attorney-led tool building. A tiered model, Explorer, Builder, and Deployer, gives qualifying attorneys access to enterprise coding tools (Claude Code and Cursor for the production tools, Microsoft Copilot Studio as the on-ramp for attorneys with no terminal experience, Lovable and Replit for low-code prototyping), per-project LLM API access, a project registry, a monthly Demo Day, and a path from MVP to production. The legal industry is in a moment the trade press calls “agentic coding”: attorneys describing tools in natural language and having AI write the code. SLW’s attorneys, most with significant engineering and scientific backgrounds before law, are uniquely positioned to lead in that movement, and Sandbox is how SLW gives them the infrastructure, the community, and the recognition framework to do it.

What Labs Has Shipped, What Is In Development, and What Is In Planning

ARTY is our proprietary AI-powered patent drafting and review platform. It is in production today and heavily used by attorneys across the firm. ARTY’s distinctive feature is not that it can draft text, as many tools can draft text. Instead, ARTY distinguishes itself by providing quality validation that runs alongside the drafting workflow. Antecedent basis, claim numbering and dependency, citation and figure cross-references, written-description support, the things that traditionally get caught (or missed) at the end of a draft pass, are surfaced as the work happens. ARTY’s Enforcement Readiness Checker runs four passes assessing claim construction, validity, and enforceability before an application is filed. AI edits are rendered as a tracked changes, requiring attorney accept-or-reject. No AI-generated content reaches a final draft without explicit attorney approval.

Alongside ARTY, the firm continues its strategic partnership with Black Hills AI (BHAI), through OTTO Hub and OTTO IP.  OTTO Hub is a proprietary data source of patent data generated through the docketing work that BHAI does for SLW. This represents a unique source of very rich data not available through other sources. We will say more about the partnership and what it brings to our practice in a later piece in this series.

In development with working prototypes: SimProf, a simulation environment for training attorneys in prosecution judgment through realistic scenarios; and SIDA, a tool that streamlines the invention disclosure process and may eventually become part of a broader client portfolio platform.

In planning: SAGE, a Strategic Advisory and Guidance Engine that will inject portfolio-level context into the prosecution workflow regardless of which AI platform an attorney is using; ATLAS, a client portfolio management portal that will give in-house counsel direct visibility into their prosecution work; and SILO, an intelligent document organizer being developed in-house for SLW’s specific document needs.

How We Build, Who Builds, and What We Will Not Compromise On

Andre L. Marais, who leads AI adoption at the firm from our Silicon Valley office, sets the strategic direction for ARTY and the broader Labs program. The ARTY team includes Steve Lundberg, Tyler Nasiedlak, Garth Vivier, and Suneel Aurora. SimProf development is led by Andre alongside Maureen Kinsler and Piers Blewett. SILO is led by Nathan Elder. The IDS Reference Review Tool is led by Bill Kalweit and Lucas Hjelle. Lucas Hjelle and Nathan Elder co-coordinate the Labs program across products.

Two functions are non-negotiable across every Labs product. Anup Suresh, who heads SLW’s security and compliance program, owns the security architecture. Tom Ernster’s IT group owns the infrastructure layer the platform runs on. Their review is required before any product ships, and before any external description of how the platform works is published.

“The Labs portfolio is alive, not fixed. Tools graduate from planning to development to production. Some prototypes get folded into other products. The discipline is keeping the standards uniform across everything we ship while letting builders move fast.”
– Nathan Elder, SLW Labs

Hard rules apply uniformly. No client data enters public AI models. Every AI-generated work product receives substantive attorney review before filing or delivery to clients. Attorneys remain personally responsible for all work product under Rule 1.1 and 37 CFR 1.56. Audit logs cover every document access, every AI interaction, every prompt, for every matter, with multi-year retention. These are not policies the firm wrote in response to client questions. They are architectural choices made before the platform existed.

Why We Are Publishing This

Two reasons. The first is that, as expected, our clients are asking what we have done to adopt AI. Corporate IP departments are evaluating outside counsel partly through an AI lens, and they deserve a clear, substantive answer rather than a marketing slogan. The second is that the patent attorneys, agents, and technical specialists we want to recruit deserve to know the drafting and editing environment in which they will be working. The traditional apprenticeship path in patent prosecution is being reshaped by AI. SLW’s answer to that reshaping is not to lower the bar; it is to use tools like SimProf to raise the floor, giving new attorneys an advantage that was previously unavailable.

“AI commoditizes execution and amplifies the value of human judgment. The firms that will lead the next decade of patent prosecution will be the ones that built rather than borrowed. SLW Labs is what that looks like at our firm.”
– Andre L. Marais, AI Adoption Lead

Over the coming weeks, SLW will publish a series of articles going deeper into each Labs product. We will start with ARTY’s quality and validation architecture, move through ARTY’s security posture, then turn to SimProf and the firm’s broader training apparatus, and close with SILO and a forward look at SAGE and ATLAS.

We will also continue to be candid about what we have learned. Some of what we tried did not work. Some of what we are building is harder than we expected. And some of it is genuinely changing how the firm practices. That is the work.