Reed Smith's AI Innovation Bet: GravityStack and Conversational Matter Intake

TL;DR#

The Reed Smith AI story is one of the few in BigLaw where a firm treated legal innovation as a product, not a slogan. The firm — ranked 37th on The American Lawyer's 2025 Am Law 200 with roughly $1.5 billion in revenue, about 1,500 lawyers, and more than 30 offices worldwide — spun out a wholly owned legal-technology subsidiary, GravityStack (now branded Gravity Stack), in 2018, becoming the first Am Law firm to commercialize its in-house tech team. GravityStack builds data products like Periscope (an e-discovery KPI platform) and, in 2024, rebuilt its entire service line around AI. Inside the firm, more than 750 lawyers and staff have completed AI certification on Harvey, which Reed Smith calls one of its most successful technology adoptions ever. Yet the firm's AI story still has one analog seam: client and matter intake, where new-business questionnaires, conflict checks, and email threads capture far less context than a lawyer needs. The next step for firms this sophisticated is conversational matter intake — an AI interview that captures the client's situation, urgency, and constraints before a partner is ever looped in.

Who Reed Smith Is#

Reed Smith is a global law firm headquartered in Pittsburgh that ranks among the largest in the world by revenue and headcount. The firm reported gross revenue of roughly $1.5 billion in 2024 and placed 37th on The American Lawyer's 2025 Am Law 200, with approximately 1,500 attorneys spread across 20 U.S. offices, 13 international offices, and dedicated Reed Smith Global Solutions delivery centers. Its practice spans litigation, regulatory work, financial services, life sciences, energy, and entertainment and media.

What sets Reed Smith apart is not size — plenty of firms are bigger — but how deliberately it has built an innovation function. Rather than treating technology as an internal cost center, the firm structured it as a business capable of serving clients directly. That posture is what makes Reed Smith a useful case study for any firm trying to figure out where artificial intelligence actually belongs in a law practice, a question also playing out at peers like Baker McKenzie, the world's largest firm by headcount, and at Kirkland & Ellis, the $7B Am Law revenue leader.

GravityStack: Productizing Legal Innovation#

GravityStack productized Reed Smith's internal legal-technology team into a standalone subsidiary that sells consulting, managed services, and data products to clients. Launched in April 2018 and led by managing director Bryon Bratcher, GravityStack was, as Legal IT Insider reported at the time, the first Am Law firm to commercialize its in-house technology group as a distinct company. Then–global managing partner Sandy Thomas framed it as "the next stage of evolution for Reed Smith's in-house legal technology team."

The subsidiary's flagship product, Periscope, is a KPI-metrics platform that pulls data from Relativity and other systems to give litigation teams real-time visibility into e-discovery costs, review productivity, and quality. Beyond products, GravityStack offers technology consulting, managed e-discovery, contract support, and data-science deployments. As Artificial Lawyer detailed in its profile of the venture, the strategic idea was to give clients a "one-stop shop" that combines a law firm, a consulting group, and a technology company.

In 2024, GravityStack went further and rebuilt its brand around AI, declaring that "there is not a single service or part of our messaging that doesn't involve AI." The firm began partnering with the AI development platform Vellum to test models and prompts — evaluating Google, Anthropic's Claude, OpenAI's GPT, and open-source models — in a sandbox before deploying them on client work. That is the same instinct that drives modern legal-tech vendors, from practice-management platforms like Clio, which is reinventing legal client intake, to specialist intake tools.

Reed Smith AI Adoption Across the Practice#

Reed Smith AI adoption inside the firm has centered on a firmwide rollout of generative-AI tooling backed by structured training and governance. According to the firm, more than 750 lawyers and professional staff have completed AI certification on Harvey, the legal-AI platform Reed Smith describes internally as one of its most successful technology adoptions. The firm supports adoption with self-service training, a dedicated AI intranet, a rapid-response internal email alias, an internal AI user community, and weekly AI office hours — a "persistent, multi-pronged" change-management approach rather than a one-time software purchase.

The firm has also invested in leadership education. In 2026 Reed Smith launched an AI Leadership Program with Cornell University, based at the Cornell Tech campus in New York, to move partners from foundational AI literacy toward applied strategy, ethics, and governance. Earlier, in 2022, it stood up a global Innovation Lab run by a business anthropologist to study how legal work actually gets done and where technology can help.

Where is the AI being pointed today? Largely at the work product: document review, e-discovery, due diligence, contract analysis, and legal research — the same use cases that dominate how BigLaw is deploying generative AI at firms like Latham & Watkins and inform the broader data-backed shifts in law-firm AI adoption. Those are high-value, high-volume tasks, and automating them frees expensive lawyer hours. But notice what this list has in common: it all happens after a matter already exists.

The Intake Gap Even Innovative Firms Have#

The gap in even the most innovative firms is at the front door: client and matter intake still runs on forms, PDFs, and email. A prospective client fills out a web contact form or a static new-matter questionnaire, an assistant re-keys it, a conflicts check runs, and eventually a lawyer reads a thin summary and calls the client back to ask the questions the form never captured. The firm has world-class AI applied to discovery and drafting, and a 2015-era workflow applied to the very first client interaction.

This matters more than it looks. Intake is where the firm learns why the client is calling, how urgent the problem is, what's already gone wrong, and what a good outcome looks like — the messy context that determines whether a matter is worth taking and how to staff it. Static forms flatten all of that into checkboxes and a 200-character "describe your issue" box. The result is the same failure mode we document in why conversational AI intake converts where legal-intake software falls short: qualified prospects abandon the form, and the ones who finish hand over data that's too shallow to act on. It's the exact problem behind the push to replace PDF intake forms with AI conversations.

Firms that have started closing this gap — like Cooley, which is replacing intake forms for its startup clients, and Wilson Sonsini, which rebuilt founder intake — are finding that the front-door conversation is as valuable as the back-office automation.

Conversational Matter Intake as the Next Step#

Conversational matter intake replaces the static new-business form with an AI-led interview that adapts to the client's answers, probes for context, and hands the lawyer a structured brief. Instead of a fixed questionnaire, an AI interviewer or a concierge agent that replaces the form asks the client to describe the situation in their own words, then follows up on anything vague — "You mentioned a dispute with a vendor; when did the contract start, and is litigation already threatened?" — the way a good intake attorney would.

The difference is not cosmetic. Here is how the two approaches compare on the dimensions that determine whether a matter gets qualified and staffed correctly:

This is the same shift specialty carriers and other regulated businesses are making. Coalition's conversational security assessment for cyber insurance and Kin Insurance's conversational property interview both replace static risk questionnaires with adaptive conversations for exactly the same reason a law firm should: the highest-value context lives in the follow-up question, not the form field. For a firm like Reed Smith that has already institutionalized innovation through GravityStack, adding an intelligent intake layer is a natural extension of the same playbook — capture the context conversationally, then feed it into the AI tooling the firm already runs. The design principles are the ones in how to design a client intake process that doesn't lose clients and the move from PDF forms to conversational triage.

Lessons for Firms Building Innovation Functions#

The main lesson from Reed Smith is that an innovation function only pays off when it reaches the client-facing front door, not just internal efficiency. Three takeaways stand out for firms building or scaling their own innovation teams:

1. Productize, don't just pilot. GravityStack worked because Reed Smith treated technology as a durable capability with an owner, a P&L, and a mandate — not a rotating set of pilots. Whether or not you spin up a subsidiary, give innovation a real home.
2. Pair tooling with change management. Reed Smith's 750-plus Harvey certifications came from office hours, an internal community, and persistent training — not from buying licenses and hoping. Adoption is a program, not a purchase, a pattern echoed in Skadden's conversational client-discovery rollout.
3. Fix the front door too. The clearest remaining ROI for most firms isn't a tenth back-office tool — it's the intake conversation, where better context improves qualification, conflicts, staffing, and conversion all at once.

Reed Smith proved a large firm can build AI capability at scale. The firms that win the next phase will point that capability at the moment the client first reaches out.

Frequently Asked Questions#

What is GravityStack?#

GravityStack is Reed Smith's wholly owned legal-technology subsidiary, launched in April 2018 and led by managing director Bryon Bratcher. It commercializes the firm's in-house technology team, selling consulting, managed e-discovery, data science, and products such as the Periscope e-discovery KPI platform to corporate legal departments and other law firms. In 2024 it rebranded around AI, rebuilding its services and messaging so that every offering incorporates artificial intelligence.

How does Reed Smith use AI?#

Reed Smith uses AI primarily for document review, e-discovery, due diligence, contract analysis, and legal research, delivered through the Harvey legal-AI platform. More than 750 of the firm's lawyers and professional staff have completed AI certification on Harvey, supported by self-service training, a dedicated AI intranet, an internal user community, and weekly AI office hours. The firm also runs a global Innovation Lab and, in 2026, launched an AI Leadership Program with Cornell University.

How big is Reed Smith?#

Reed Smith is a global law firm that reported roughly $1.5 billion in gross revenue in 2024 and ranked 37th on The American Lawyer's 2025 Am Law 200. It employs approximately 1,500 attorneys across 20 U.S. offices, 13 international offices, and dedicated Reed Smith Global Solutions delivery centers, with headquarters in Pittsburgh.

What is conversational matter intake?#

Conversational matter intake is an AI-led interview that replaces static new-business forms and questionnaires at a law firm's front door. Instead of asking a prospective client to fill in fixed fields, an AI interviewer asks them to describe their situation in their own words and follows up on vague or incomplete answers, then delivers a structured, context-rich brief to the intake lawyer. It captures urgency, constraints, and the "why now" that forms miss.

Why do innovative law firms still struggle with intake?#

Innovative law firms still struggle with intake because their AI investments have concentrated on post-matter work — discovery, drafting, research — while the first client interaction stayed on forms and email. Intake is harder to automate with off-the-shelf tools and is often owned by business-development or operations teams rather than the innovation group. That leaves a sophisticated firm capturing shallow, form-shaped data at the exact moment context matters most.

Conclusion#

Reed Smith AI innovation is one of the clearest proofs that a large firm can turn technology into a genuine capability: GravityStack productized the firm's in-house tech team, Harvey put certified AI tools in the hands of more than 750 lawyers and staff, and a Cornell partnership is training partners to lead AI strategy. The unfinished work is the front door. Even at a firm this advanced, client and matter intake still runs on forms that flatten the context lawyers most need — the same gap that reappears across BigLaw's client-intake efforts and the wider legal-intake market.

Conversational matter intake is the logical next step, and it's built for exactly this: capturing the client's situation, urgency, and constraints in their own words before a lawyer is looped in. If your firm has invested in AI everywhere except the first conversation, start a matter-intake interview with Perspective AI and see how much more context an adaptive conversation captures than any form ever will.