Sectors · Law firms

AI for small law firms, with professional secrecy designed in

Lawyers should spend their time on legal work, not on retyping, chasing and searching. I automate the admin around the legal work, on the systems you already use, and set it up so it stands up to your duty of confidentiality.

The admin around the legal work

In a small firm, much of the non-billable week goes on the same handful of jobs. These are the ones I look at first:

  • New-matter intake. Reading enquiries, collecting the parties and facts, preparing the conflict check and the engagement letter.
  • The shared inbox. Sorting correspondence from clients, opposing counsel, courts and authorities, and making sure nothing urgent waits.
  • Document assembly. Engagement letters, standard letters and recurring contract types built from templates by hand.
  • Finding the precedent. Searching folders for the last time the firm drafted a similar clause or submission.
  • Deadlines. Reading dates out of court and authority letters and getting them into the calendar.
  • Time narratives and billing preparation at the end of the month.

Workflows that suit a small law firm

  • Matter intake. An enquiry is read, the parties and key facts are pulled into your intake form, and the names for the conflict check are listed for a person to run in your practice system.
  • Inbox triage. Incoming mail is classified by matter and urgency, routine acknowledgements are drafted for review, and anything with a deadline is flagged. How AI email triage works.
  • Document assembly. First drafts of engagement letters and standard documents from your own templates and the intake data, for a lawyer to finalise.
  • Precedent and knowledge search. Ask a question in plain language and get answers from the firm's own templates and past work, with a link to the source and respecting who may see which matter.
  • Deadline capture. Dates are read from incoming court and authority letters and proposed as calendar entries. A person confirms every one.
  • Time narratives. Draft entries from the day's emails and calendar, for each lawyer to check and correct.

Professional secrecy and the nFADP

For a law firm, the bar is professional secrecy under Art. 321 of the Swiss Criminal Code and the Lawyers Act, not only data protection law. The analysis prepared for the Swiss Bar Association treats a cloud provider as an auxiliary person, so using one is possible if the firm chooses, instructs and supervises it carefully and binds it to confidentiality by contract. I cover this in detail in can a Swiss law firm use ChatGPT or Copilot with client data?

On the data protection side, the Federal Data Protection and Information Commissioner has confirmed that the revised Federal Act on Data Protection applies directly to AI. [FDPIC] Any AI service that processes client data for you is a processor: the Act allows that only by contract, only if no duty of confidentiality prohibits it, and only once you have satisfied yourself that the provider can guarantee data security. [Art. 9 FADP]

In practice, every workflow I build for a law firm has:

  • one business AI service under contract, with no training on your data, never personal accounts;
  • access that follows your matter permissions, tidied before anything is switched on;
  • a lawyer approving anything that reaches a client, a court or the other side;
  • a written one-page AI rule for the firm and a line in your engagement terms.

This is not legal advice on your professional obligations; for that, speak to your counsel or bar association.

From first call to a workflow your team runs

01

Discovery call

We talk through your firm, the bottlenecks, the tools you use and the people who will use whatever we build.

02

AI opportunity map

What can be automated, what should not be, the expected return and the quick wins, written down.

03

Build and test

The workflow or tool is built on your real processes and tested with your team's feedback.

04

Deploy and train

It goes live with your team, documented in plain English, with support while people adopt it.

For a law firm, the discovery stage includes a partner, whoever runs the office and whoever handles the inbox day to day. The audit looks at how work flows, not at client files; where a real example helps, we agree what I see and how before I see it.

Most firms start with the AI Opportunity Audit (CHF 750 to 1,500, one to two weeks). A first automated workflow starts from CHF 1,500 and a custom tool from CHF 5,000. Every price is fixed before work starts. See pricing.

Common questions

Can a small Swiss law firm use AI with client files?

Yes, with a business AI service under contract that does not train on your data, access limited to the right people, and a lawyer reviewing outputs. Not with personal accounts. Full explanation.

Will AI give legal advice to our clients?

No. The workflows I build handle admin around the legal work: sorting, drafting from your templates, finding precedents and capturing dates. A lawyer approves anything that leaves the firm.

Do you work with our practice management software?

Where it can exchange data, through an interface or exports, workflows can connect to it. Where it cannot, the workflow works on email and documents and a person carries the result across. The audit establishes which applies.

What does it cost for a law firm?

The same published prices as any firm: the audit is CHF 750 to 1,500, a first workflow from CHF 1,500, a custom tool from CHF 5,000, each fixed before work starts.

Is Microsoft 365 Copilot a good starting point for a law firm?

Often, if the firm already runs Microsoft 365, because it works inside the tenant you already trust. Fix over-shared folders first: Copilot shows people everything they already have access to.

Want to know where AI could save your firm time?

Book a short call. We will identify the first two or three workflows worth automating. No deck, no jargon, no hard sell.

Book an AI opportunity call