Legal

Streamlining invoicing and administrative legal work

AI can reduce time spent on repetitive administrative work such as invoicing and other operational tasks around a legal practice.

Why the human is still essential here

The professional still oversees accuracy, billing appropriateness, and compliance with client and practice requirements. Administrative efficiency improves, but accountability stays with the human practitioner.

How people use this

Draft invoice narratives

AI converts time entries, notes, or calendar activity into clearer billing descriptions that the lawyer reviews before sending invoices.

Clio Manage AI / Billables AI

Missing-time detection

AI surfaces likely billable work that was never entered, helping the practice capture time more consistently without rebuilding invoices by hand.

Billables AI / Clio Manage AI

Administrative follow-up drafts

AI drafts payment reminders, invoice explanations, and other routine billing communications so staff can send polished messages faster.

Clio Manage AI / MyCase

Need Help Implementing AI in Your Organization?

I help companies navigate AI adoption -- from strategy to production. Whether you are building your first LLM-powered feature or scaling an agentic system, I can help you get it right.

LLM Orchestration

Design and build LLM-powered products and agentic systems

AI Strategy

Go from idea to production with a clear implementation roadmap

Compliance & Safety

Build AI with human-in-the-loop in regulated environments

Related Prompts (2)

Latest community stories (1)

Personal Story
LinkedIn

I use AI for almost everything in my practice.

I use AI for almost everything in my practice. Except the mediation itself.

Last week I joined the Claude Code Hackathon at Baxter Building, organized by Recht in je Oor and Guldemond Advocaten . There were 30 of us, lawyers and jurists, working in five groups on prototypes. Most of us had never really coded before.


What the hackathon clarified


The speed at which you could build what you wanted, if you sat down and explained it properly, was striking. But the more interesting thing was the pattern that every prototype attacked the same chokepoints: the time-consuming, repetitive parts of legal work. Nobody built a generic legal AI. Everyone built something tailor-made for a specific friction in their domain.


How I use AI in my own practice


I use AI for almost everything around my mediation work. Writing, positioning, marketing, strategizing, invoicing, structuring my thinking when I am stuck. What I deliberately keep out is the conversational core. The intakes, the plenary sessions, the calls in between, the back-and-forth about who sits where and when we meet. That is not an efficiency problem waiting to be solved. That is the work itself.


Why the conversation is not a chokepoint


Parties almost never arrive knowing what they want. They arrive feeling. Angry, misunderstood, stuck in a story they have been telling themselves for months. The job is to help them get underneath that, and you cannot do it with a questionnaire or a summarized position paper. It happens through conversation, often the same one more than once, until a new layer surfaces. A messy process, with steps forwards and leaps back, tensions, outbursts and reconciliations.


The same applies to interactions that look like overhead. A client calls to add something quickly and it is never quick, because if they felt the need to pick up the phone, something is bothering them. The scheduling, the composition of the table, all of it carries information and builds trust. You can automate it away. You will be faster. You will also be working with less.


Efficiency, effectiveness, sustainability


Mediation is faster than arbitration or litigation. But speed is not the point. The point is that the outcome holds. A compromise that surfaces from one conversation and a processed position paper might look reasonable and still fall apart within months, because the parties never got to what was boiling underneath. The slower path is what makes it sustainable. Go slow to go fast.


There a striking meta-phore about all this: The same patience that makes a mediation hold is the patience that makes AI actually useful. If you give it a three-line prompt for something complex, it will not understand you. You have to sit down with it. Go back and forth. Say the same thing in different words until it clicks. The three-line prompt and the five-minute intake are the same mistake.


I'm a big believer that AI belongs in the chokepoints. The conversation between the parties is not one of them.

EG
Ewout GoudsmedtMediator and negotiation advisor
May 12, 2026