Best Practice

Dr Heidi K. Gardner is a Distinguished Fellow at Harvard Law School and the author of Smart Collaboration and co-author of Smarter Collaboration. An organisational behaviour researcher and former McKinsey consultant, she has spent years studying how professionals share expertise, build trust and work together to solve complex problems. Her work explores a question that is becoming more pressing as AI changes legal work: how do you turn a group of talented individuals into a team that delivers better outcomes for clients?In this conversation, we explore what happens when lawyers ask AI questions they would previously have taken to a colleague, why bringing people back to the office doesn’t automatically create collaboration, and how junior lawyers can keep learning from the people around them.We discuss:* What happens when lawyers turn to AI agents instead of colleagues.* Why office attendance alone doesn’t create collaboration.* What juniors learn in the three minutes before and after a client call.* Why the billable hour can discourage teamwork.* How trust and shared goals help law firms collaborate better.Where to find Heidi:Website: smartercollaboration.comHeidi’s LinkedInBooks: Smart Collaboration My biggest takeaways from this conversation:1. Getting an answer isn’t the same as learning from someone.I’m guilty of turning to AI before asking a colleague. It’s convenient, but this conversation made me think more carefully about what that interaction would otherwise have given me. A colleague might challenge my assumptions, explain the context or spot that I’m asking the wrong question. As a junior, I want to keep making room for those conversations alongside using AI.2. The office is only valuable if we make use of being together.Heidi challenged the idea that attendance alone creates collaboration. If senior people sit behind closed doors and juniors spend the day on Teams, being in the same building achieves much less than it could. That also made me reflect on my own habits: when there’s a meeting room booked, joining from my desk means missing an opportunity to sit with the team and talk afterwards.3. A three-minute debrief can be a powerful training tool.At McKinsey, Heidi would be invited to listen to a partner’s client call and then discuss it immediately afterwards. What was the client really asking? What did the partner miss? What should happen next? Knowing those questions were coming made her an active listener. I liked how practical this was: a short conversation can help a junior understand decisions they might otherwise just observe.4. “I don’t know, but I know someone who does” is a valuable answer.It’s easy to feel that credibility comes from always knowing the answer. Heidi’s point was that clients recognise the limits of any individual’s expertise. Being clear about those limits, and having trusted people you can bring in, can build confidence. For me, that makes getting to know colleagues across the firm part of becoming a better adviser.5. Collaboration depends on the behaviour we reward.If a firm celebrates individual billing and client wins, it shouldn’t be surprising when people prioritise those things. Heidi’s answer to designing a firm from scratch started with shared goals, developing people and valuing different strengths. It left me thinking that collaboration needs to show up in everyday decisions: who gets recognised, who gets promoted and whether helping someone else succeed counts as success too.This post has bonus content for paid subscribers. Upgrade to get full access. Get full access to Best Practice at bestpracticeai.substack.com/subscribe

Bernadette Bulacan is Chief Evangelist at Icertis and has spent more than 20 years working at the intersection of law and technology. Through her work with some of the world’s largest businesses such as HP and BAE Systems, she sees how legal teams are changing the way they negotiate, manage and learn from contracts. In our conversation, we get into:* Why AI represents a bigger shift for legal work than previous generations of technology* Where contracting actually slows down, and why legal isn’t always the bottleneck* What AI agents need before they can negotiate contracts autonomously* What human-centred contracting looks like, and where legal engineers fit in* Whether AI will give lawyers time back or leave them busier than ever* What junior lawyers should learn now, and how to use AI without surrendering their judgmentWhere to find Bernadette:• LinkedIn• Icertis: icertis.comExplore more from Best Practice:Legal AI news, interviews and commentary at bestpractice.media.My thoughts from my conversation with BernadetteMy biggest takeaway is that agent-to-agent negotiation will depend on how well a business can explain its own position. An AI agent needs to know more than what a clause means: it needs to understand what the company has accepted before, where it can compromise and when to bring a lawyer in. That makes the work of organising precedents, updating playbooks and setting clear boundaries much more important. Otherwise, we risk making poor decisions faster.I also came away thinking about what this means for junior lawyers like me. If agents handle more of the drafting and negotiation, understanding why a business wants a particular outcome becomes even more valuable. Bernadette’s advice to challenge AI as you would a colleague really stuck with me. I think its becoming important to use these tools to explore decisions and develop my judgment, but also whilst making sure I can still explain and stand behind the advice I give.This post has bonus content for paid subscribers. Upgrade to get full access. Get full access to Best Practice at bestpracticeai.substack.com/subscribe

“Humans are now the bottleneck.”I spoke with Rafie Faruq, co-founder and CEO of Genie AI, about what happens to legal work when AI can increasingly do the work itself.Rafie has been building with generative AI since long before ChatGPT, founding Genie AI in 2017 after studying machine learning at UCL.We discussed:→ Why coming into legal tech without being a lawyer gave Rafie a different perspective on redesigning legal workflows, rather than simply making existing ones faster.→ Why the lawyer of the future could look more like a “trust architect”, setting policies, preferences and risk tolerances for AI agents rather than manually doing every task.→ Why accuracy may become less of a differentiator for AI, while consistency, judgment, taste and organisational preferences become much more important.→ Whether AI will actually give us more free time — or simply create even more demand for work as productivity increases.We also spoke about AI anxiety, agent-to-agent negotiation, why future AI may communicate in ways designed for machines rather than humans, and Rafie’s advice for young people entering a world where the technology is changing incredibly quickly.One of my favourite takeaways was his advice not to constantly chase whatever happens to be hot right now, but to think from first principles and build conviction around ideas you genuinely believe in.I hope you enjoy listening just as much as I did recording it!Want to feature in the Best Practice podcast? Email me at [email protected]:00:00 Rafie’s background and building Genie AI04:01 Generative AI before ChatGPT08:40 What skills will future lawyers need?11:30 Agent-to-agent negotiation14:17 Why humans are becoming the bottleneck17:35 AI anxiety and endless productivity21:02 Why taste and consistency matter27:17 What happens to law firms when clients have AI?30:27 Rafie’s advice for the next generation Get full access to Best Practice at bestpracticeai.substack.com/subscribe

* Guest: Will Chen* Role: Founder, Mike OSS & LawProf* Website: mikeoss.comThis episode is brought to you by Aline. Try Aline’s trusted and verified legal AI today for free.I spoke with Will Chen, a former Latham & Watkins lawyer who is now building Mike OSS, an open-source legal AI platform, alongside LawProf, the legal education platform he originally built while at university.Will has become one of the most-followed voices in the Legal AI space; particularly around open source, the economics of legal AI platforms and whether law firms should be building more technology themselves.We discussed:→ Why Will decided to open source Mike OSS, and whether open-source legal AI can become commercially sustainable.→ Whether legal AI platforms are really differentiated products or, in Will’s words, increasingly competing on deployment, distribution and access to models.→ Why law firms may need to start hiring and paying top engineering talent very differently if they want to build serious technology internally.→ The rise of open-weight and Chinese AI models, and what a rapidly moving model frontier means for companies trying to post-train their own legal models.→ Will’s advice for junior lawyers entering the profession as AI becomes significantly more capable.We also spoke about vibe coding, why lawyers still need strong fundamentals even as AI improves.I hope you enjoy listening just as much as I did recording it!Want to feature in the Best Practice podcast? Email me at [email protected]:01:03 Introduction and Will’s background in law and AI03:34 Why Open Source Legal AI Matters08:17 Coding, AI, and Legal Judgment14:17 The Value of Legal AI Vendors22:47 Build vs. Buy for Law Firms29:10 The Rise of Chinese AI Models33:02 Career Advice for Future Lawyers Get full access to Best Practice at bestpracticeai.substack.com/subscribe

* Guest: Kenzo Onumonu* Role: Founder, Correm * Website: correm.aiThis episode is brought to you by Aline. Try Aline's trusted and verified legal AI today for free.I spoke with Kenzo, a former Bird & Bird lawyer and in-house GC who is now building Correm, a platform for companies to access and manage fractional legal talent.Having sat on both sides of the legal market, Kenzo has seen first-hand how difficult it can be for businesses to get the right legal support at the right time.Correm is built around a different model, combining flexible legal talent with technology, automated billing and AI tools to make it easier for companies to build legal capacity around the work they actually have.We discussed:→ Why fractional legal teams are becoming a viable alternative to traditional hiring.→ What Kenzo learned going from private practice to becoming a GC, and then a founder.→ How AI changes the economics of flexible legal work and what the legal team of the future could look like.→ Why lawyers may increasingly need to think about their careers as a collection of skills and experiences rather than one linear path.→ Kenzo’s tips for junior lawyers navigating their career in an AI landscape.I hope you enjoy listening just as much as I did recording it!Want to feature in the Best Practice podcast? Email me at [email protected] Get full access to Best Practice at bestpracticeai.substack.com/subscribe

Guest: Jake JonesRole: Co-founder and COO, FlankLinkedIn: https://www.linkedin.com/in/jake-jones-b8769b75/Website: flank.aiThis episode is brought to you by Aline. Try Aline's trusted and verified legal AI today for free.Sources mentioned in the episode:* Flank — flank.ai* Insights portal — insights.flank.ai* The Intake (Substack, Flank's team blog) — intakezero.substack.comI spoke with Jake Jones, co-founder and COO of Flank. He’s spent years at the forefront of legal AI across what he counts as four different paradigms and half a dozen hype cycles. He started the business in 2018 as Legal OS, modelling how experienced lawyers make decisions as knowledge graphs, built on his co-founder's father's academic work, and sold it to German law firms doing M&A. In 2021 he moved the business in-house, rebuilt it on LLMs, and now Flank runs agentic legal work inside enterprise legal teams under a model where they don't get paid until the outcome is live.We discussed:→ Why vendors are already moving beyond "agentic AI" to full end to end outcomes.→ The difference between selling software and selling an outcome, and why so many legal AI vendors are stuck between the two.→ His theory that legal is heading toward the same moment retail banking hit with digital challenger banks in 2015, and what Big Law should do about it before it's too late.I hope you enjoy listening just as much as I did recording it!Want to feature in the Best Practice podcast? Email [email protected] Get full access to Best Practice at bestpracticeai.substack.com/subscribe

Guest: Brent FareseRole: CEO and Co-Founder, AlineWebsite: aline.coThis episode is brought to you by Aline. Try Aline’s trusted and verified legal AI today for free. I spoke with Brent Farese, CEO and co-founder of Aline, an agentic legal AI platform built exclusively for in-house teams. Brent trained as an IP attorney with a mechanical engineering and computer science background, then worked as Assistant General Counsel one of the world’s leading enterprise blockchain companies. We discussed:→ The three tiers of document search (keyword, semantic, agentic) and why getting this right is the hard engineering problem nobody talks about→ Why legal AI is roughly two years behind agentic coding — and what infrastructure still needs to be built before it catches up→ Why frontier model providers entering legal is actually a tailwind for specialist platforms→Why making legal work cheaper and faster is likely to produce more legal work, not less→ Why “building the GitHub for law”, with proper version control, is needed in 2026.I hope you enjoy listening just as much as I did recording it.Want to feature in the next Best Practice podcast? Email [email protected] Get full access to Best Practice at bestpracticeai.substack.com/subscribe

Is owning the data still the moat in legal AI?I spoke to Sean Fitzpatrick, a senior leader at LexisNexis - one of the largest legal AI companies in the world.Sean Fitzpatrick is CEO of the Global Legal Business at LexisNexis. He has been there for 21 years, growing up through the shifts from digital to cloud to AI, after starting out as a consultant at McKinsey.We discussed:→ Whether 200 billion proprietary documents are still a defining asset - or whether the nature of the moat has changed.→ Why "probably right" is the standard for frontier models, and why that standard fails in law.→ How their partnership with Harvey actually works.→ Sean's thoughts on hallucinations in legal AI.→ What happens to the junior lawyer apprenticeship model when the grunt work that builds judgment disappears.Sean has watched every major technology shift in legal over two decades, and makes the case for why content plus technology is far harder to displace than either alone. It was an honour to speak to Sean, and I hope you take something useful away from this.Know someone, or want to feature on the Best Practise Podcast? Email [email protected] Get full access to Best Practice at bestpracticeai.substack.com/subscribe

Guest: Andrew Thompson Role: CTO, Orbital LinkedIn: https://www.linkedin.com/in/andrewmatthewthompson/ Website: www.orbital.tech (company) | www.andrewthompson.co (personal blog) This AI platform powered over 200,000 property transactions last year and raised $60M in January to expand even further.I spoke with Andrew Thompson, CTO at Orbital. Andrew previously built the Caffè Nero app, ran product and tech at Appear Here, founded his own AI company, and now leads engineering, product and a team of in-house real estate lawyers at Orbital.We discussed:→ How Orbital mix both generalist LLM’s and their own classical ML into one product.→ Why legal is roughly 12-18 months behind software engineering, and how to use that lag.→ What “ship early and often” actually looks like inside a regulated industry.→ Andrew’s tips and advice for Andrew is an absolute genius in everything LLM’s and how to make the most of it in the physical space around us. I really hope you take away something useful from this.Want to feature in the Best Practice podcast? Email [email protected] mentioned in the episode:* Orbital’s $60M Series B (Jan 2026, led by Brighton Park Capital): https://tech.orbitalwitness.com/posts/2026-01-23-series-b-announcement/* Caffè Nero (Andrew’s previous build): https://caffenero.com* Appear Here (Andrew’s previous build): https://www.appearhere.co.uk* Seedcamp (where Andrew was a startup mentor): https://seedcamp.com* ElevenLabs (referenced re: agentic insurance): https://elevenlabs.io* Legalweek New York 2026 (referenced throughout)* McKinsey research on automatable legal work (referenced)* OpenAI ChatGPT and GPT-4 (referenced as the inflection point)* Anthropic Claude (referenced)* Google BERT and T5 (Orbital’s pre-LLM stack) Get full access to Best Practice at bestpracticeai.substack.com/subscribe

“Crack the chronology, crack the case.”That’s what Gregory Mostyn‘s father - a barrister, then judge for 42 years - was told on his first day in chambers.Decades later, it’s the founding insight behind Wexler.I spoke with Gregory Mostyn, co-founder and CEO of Wexler, the fact intelligence platform now used by huge massive circle firms in their litigation teams.Greg isn’t a lawyer. He grew up watching his father - who handled some of the highest-profile family cases in England - come home with ten ring binders to read before court the next morning. That image, plus a stint at Entrepreneur First, became Wexler. The company has 20x’d ARR since pre-seed and just raised $5.3m led by Pear VC.We discussed:→ Why Wexler went deep on litigation when every other legal AI startup was going broad→ The “verification tax” and where fully agentic legal AI actually breaks down→ Why consumption pricing beats per-seat for litigation work (and how they structure it)→ How his 67-year-old father is now vibe-coding apps for his tennis club→ What it takes to close a Big Law deal in a weekI absolutely loved recording this episode and I hope you enjoy listening to it just as much as I did.Know someone that might want to come on the Best Practice Podcast? Email me at [email protected] Get full access to Best Practice at bestpracticeai.substack.com/subscribe