MILabs
Integrations

Smokeball AI: where AI helps around legal matters and documents

What AI can do inside and around Smokeball for a law firm: time capture, document review, billing follow-up and email triage.

Updated 2026-06-22

Smokeball: what AI can do with it

Smokeball holds the working record of a law firm: matters, contacts, time, documents and the email that goes with them. AI can help at the edges of that record rather than replacing it. A lawyer or clerk still decides what goes to a client, what is billed and what is filed, but a well-scoped assistant can read a long document and answer questions about it, sort incoming email by matter, or prepare a first draft of a letter from details already in the file. Each of these depends on the data being tidy, so the first useful step is usually a review of how time, contacts and precedents are recorded. From there, AI can take on repetitive chasing and recording work, such as proposing time entries for review or drafting reminders for overdue invoices. For a firm already using Smokeball, the value comes from shortening the distance between the matter file and the next action. This page covers where those gains are realistic, what Smokeball supports, and where human review must stay.

Smokeball in plain language

A matter is the unit of work in a law firm: one client problem, such as a property purchase, a dispute or a family separation, together with everything that gathers around it. Software that organises matters tries to hold that material in one place, so that a fee earner opening the file can see what has been agreed, what has been sent and what is still owed. As matters accumulate, a firm builds a large store of routine text, timings and correspondence that follows recognisable patterns.

Artificial intelligence, in this setting, usually means language models that can read and write text, plus automation tools that move information between systems on a schedule or when something happens. These sit beside the records rather than inside them. A language model can summarise a document, but it has no knowledge of the firm's billing rules or of which matters are confidential unless someone supplies that context on purpose. Automation tools handle the plumbing, such as noticing a new calendar entry and passing its details along.

The useful question is therefore less about which product to buy and more about which steps in the file repeat, follow a predictable pattern and can be checked afterwards. Those steps are where a model or workflow can prepare a first draft, while a lawyer keeps the judgement.

What Smokeball does

Smokeball is legal practice management software built for law firms. It keeps each matter in one place, linking contacts, documents, email, time records and billing to the client file. Fee earners use it to record time, draft documents from templates and keep correspondence attached to the right matter. Office managers use the same records for billing and reporting. It is used by law firms in Australia and the United Kingdom, and it suits firms that want the matter file to be the main record. Because the matter record holds much of the context, it is a natural place to look for steps that repeat across many matters.

Where AI helps

  • Drafting first versions of routine letters and emails from matter details, for a lawyer to review before anything is sent.
  • Summarising long documents on a matter file and answering questions about them, with pointers back to the source page.
  • Sorting incoming email by matter, urgency and the action it needs, so the responsible person sees it first.
  • Proposing time entries from calendar events and correspondence, which a fee earner checks and approves.
  • Preparing reminders for overdue invoices from the billing record, held for staff to approve before they go out.

Easy wins around it

  • Your matter documents and emails already give AI plenty to summarise, draft and tidy, with no direct data connection needed to begin.
  • Any AI draft counts as a starting point. A lawyer remains responsible for what the client receives and for what is billed.
  • Client confidentiality and privilege apply to anything sent to an AI model, so hosting, retention and data terms should be settled before any matter content is used.
  • Trust and regulated accounting records are outside this page's scope and should stay under the firm's existing controls.

How AI connects to it

  • Smokeball has its own connections to Microsoft Outlook and Word, which are the usual places where matter email and documents enter the file. That makes your inbox and your documents natural places for AI to help: summarising a long email chain, drafting a letter or tidying precedent wording.

    Source: smokeball.com

Smokeball in practice

In a working firm, this work usually starts with a trigger that people already recognise. A new matter is opened, a document lands in the file, or a billing period closes. An automation notices the event, gathers the details the firm already holds, and places a draft where the responsible person will see it. The draft might be a list of proposed time entries, a short note on what a long agreement says about termination, or a summary of a settlement letter.

Most of the effort goes into the handoff rather than the generation. Someone has to decide who reviews each kind of output, how long a draft can wait before it is followed up, and what happens when the assistant is unsure. Firms often assign one reviewer per workflow so responsibility stays clear, and they keep a short log of rejected drafts, since those rejections show where instructions or source data need tightening.

Access is a quieter part of the work. The workflow needs a defined user or service account with limited rights, and someone must check which folders and mailboxes it can read. Each change to a template, a fee rate or a precedent clause can alter the output, so the reviewer should hear about those changes as they happen.

good questions, straight answers

Frequently asked questions

Which tasks in Smokeball suit automation?
Repetitive steps that depend on information already held in the matter file suit automation best. Examples include sorting incoming email to the right matter, proposing time entries for approval, and preparing reminders for overdue invoices. Tasks that need a legal judgement, such as deciding what advice to give, should stay with a lawyer, with automation handling the groundwork around them.
Can an AI assistant write to Smokeball?
That depends on the connection and the edition, which is why it should be checked before design starts. An assistant can prepare drafts, time entries or file notes for a person to review, and a staff member then saves them into the matter. Writing directly into records without review is possible in principle but adds risk, so a human approval step is the sensible default.
Who looks after the connection to Smokeball?
A named person inside the firm should own it, usually someone in operations or the office manager role who knows how matters are opened and billed. An outside partner can build and document the workflow, and MILabs, a Sydney agency, can deliver that work remotely outside Sydney. Ownership matters because Smokeball settings, precedents and staff roles change over time, and the connection needs someone to notice when they do.
Is AI worth it for a small Smokeball team?
Yes, if the team has a clear, repeated task that already leaves a record in the file. A small team usually gets more from one well-chosen workflow, reviewed carefully, than from a broad rollout. This page cannot promise savings, since results depend on matter volume, data quality and how the firm reviews output, so a scoping conversation is the sensible first step.
What can AI do inside Smokeball?
AI works best around Smokeball, helping with the work that surrounds its records. It can summarise documents already on a matter, draft a letter or time note, and answer questions about precedent wording. It can also turn a long email thread into a tidy file note and sketch a client update ready for your polish. The lawyer stays in charge of what goes onto the matter.
let's make something happen

Workflow automation: how we can help

Choose how you want to work on this with MILabs. Both options start with an enquiry.

Discovery is free, no obligation from the first chat.

Let us build it for you

We build it and look after it.

Build it yourself, we'll help

You build it with our direction, teaching and a weekly check-in.