Examples · Law offices

AI automation for law offices

Solo and small firms, where client files often have to stay in the office. Here is where AI could take routine work off your plate, and where you stay in charge.

These are illustrative designs, not case studies. What we build for you depends on your tools, your data and your rules.

A typical business like yours

Who
A firm of 2 to 15 people: attorneys, one or two paralegals and an office manager.
Common tools
Practice-management software such as Clio, MyCase or PracticePanther; Microsoft 365 with Outlook and Word; a document store such as NetDocuments, SharePoint or a file server; an e-signature service; an e-filing service provider for court filings.
A normal week
Inquiries arrive by phone and web form, someone runs a conflict check and sends an engagement letter, matters produce documents, deadlines and time entries, and bills go out monthly.
Where the time goes
Slow intake, deadlines copied by hand from court notices, document review that eats associate time, time entries reconstructed at month end, and client status questions.
Documents
Engagement letters, pleadings, discovery productions, contracts, correspondence and billing records, usually in the document system and in email. Confidentiality duties mean any AI tool has to be vetted first, and some firms prefer to keep processing on their own systems.

Tools named here are ones such businesses commonly use. We confirm what your own software allows before promising a connection.

Examples

Documents & complianceHybrid option

New-client intake with conflict check and engagement letter

An inquiry and its attachments become a conflict report, a matter summary and a draft engagement letter, and the documents can be read on the firm's own hardware.

Starts when

A prospective client submits the intake form with documents.

What happens

  1. Read the documents and list the parties, dates and type of matter
  2. Search every party against your client and matter history for conflicts
  3. Pull any response date stated in the papers and flag it for confirmation
  4. Draft the engagement letter from the firm's template and fee schedule

In more detail

An automation like this reads a prospective client's intake form and attached papers on the firm's own computer. It lists every person and company named, the dates and the type of matter, then searches those names against the firm's past clients, opposing parties and declined prospects and writes a conflict report showing exact and close matches. If the papers state a response date, or show when the person was served, it flags that date for a lawyer to confirm. It then drafts an engagement letter from the firm's own template and fee schedule. An attorney reviews the conflict report, decides whether to take the matter, and edits and sends the letter for signature. It needs read access to client and matter history, either through the practice-management system's API or a regular export. Fenrir Labs could design this so documents never leave the office network.

You stay in control

An attorney reviews the conflict report and signs off on the letter.

Connects to

  • Intake form
  • Practice management
  • Document storage
  • E-signature

Questions we would ask you

  • Which practice-management product and tier do you use, and do you enter opposing parties and related persons on every matter?
  • Do you keep a list of prospective clients you declined, and where?
  • How do inquiries arrive today: web form, email, phone, referral? Which intake tool, if any?
  • How many engagement-letter templates do you have, and how do fee terms vary by matter type?
  • Who decides a conflict today, and what does a "close call" look like at your firm?
Discuss this further with the Live Agent

Documents & complianceHybrid option

First-pass review of a document production

Thousands of produced pages are sorted, tagged by issue and summarized with page references, so review starts with the documents that matter.

Starts when

A production is loaded into the matter folder.

What happens

  1. Make every page searchable and group duplicates and email threads
  2. Tag each document against the issues and names you list
  3. Flag documents that may be privileged for a separate look
  4. Write a summary of each key document with citations to the page

In more detail

An automation like this takes a document production saved in a matter folder and prepares it for attorney review on the firm's own hardware. It makes every page searchable, groups duplicates and email threads, tags each document against the issues and names the attorney lists, flags documents that may be privileged for a separate look, and writes short summaries of the likely key documents with page references. The result is a spreadsheet and a suggested reading order. Attorneys make every relevance and privilege decision; the tags only decide where reading starts. It needs a capable GPU machine in the office, an issue list from the attorney, and patience for overnight processing on larger sets. For very large or contested productions, a hosted review platform is usually the better choice, and Fenrir Labs could say so.

You stay in control

Attorneys make every relevance and privilege decision; the tags are a starting order.

Connects to

  • Document storage
  • Local AI server
  • Review spreadsheet

Questions we would ask you

  • How large are your typical productions (pages, emails, native files), and how many a year?
  • In what form do they arrive: PDFs, emails, load files from a vendor?
  • Who does first-pass review today, and how long does it take on a typical production?
  • Do you use any review platform now (Logikcull, Everlaw, MyCase Discovery Assistant), and what do you dislike about it?
  • Do your protective orders restrict where produced documents can be stored or processed?
Discuss this further with the Live Agent

Scheduling & dispatchHybrid option

Court notices to calendar entries

E-filing notices are read as they arrive. Hearing dates and stated deadlines are proposed as calendar entries with the source notice attached.

Starts when

A notice arrives from the court's e-filing system.

What happens

  1. Match the notice to the matter by case number
  2. Extract the hearing date, the filing and any deadline it states
  3. Propose calendar entries and reminders for the responsible attorney
  4. File the notice in the matter folder under a consistent name

In more detail

An automation like this reads e-filing notices as they arrive, such as eFileTexas acceptance and service emails and federal court notices. It matches each one to the right matter by cause number, saves the notice and the filed document to the matter folder under a consistent name, and pulls out any hearing date or deadline the notice or attached order actually states. It proposes calendar entries and reminders for the responsible attorney, each shown beside the sentence it came from. A paralegal or attorney confirms every date before it reaches the calendar. It does not replace court-rules calendaring software for deadlines set by rule rather than stated in the notice. It needs read access to the mailbox where notices land and to the matter list in the practice-management system.

You stay in control

The attorney or paralegal confirms each date before it is added to the calendar.

Connects to

  • Email
  • Practice management
  • Calendar
  • Document storage

Questions we would ask you

  • Which courts do you file in (Harris County district courts, county courts at law, justice courts, federal Southern District), and how many notices arrive a week?
  • Where do court notices land today, and who reads them?
  • Do you use a court-rules calendaring tool, and which jurisdictions does it cover?
  • Is the cause number stored on every matter in your practice-management system?
  • Who confirms a date before it goes on the calendar, and who is the backup when they are out?
Discuss this further with the Live Agent

Finance & billing

Time entries rebuilt from the day's work

At the end of the day each timekeeper gets a draft of their entries, pieced together from calendar events, emails, calls and documents they worked on.

Starts when

End of each working day.

What happens

  1. Collect the day's meetings, calls, sent emails and edited documents
  2. Group the activity by client and matter
  3. Draft a narrative and a suggested duration for each entry
  4. List gaps in the day where no activity was captured

In more detail

An automation like this gives each timekeeper a private draft of the day's time entries every evening. It reads the day's calendar meetings, sent emails, phone calls from the phone system's log, and documents the person edited, groups that activity by client and matter, and drafts a short billing narrative and a suggested duration for each entry based on the recorded activity. It also lists gaps where nothing was captured so the timekeeper can fill them in while they still remember. Only the timekeeper sees the draft, and they edit and release their own entries; nothing reaches the billing system until they do. It needs read access to calendar, mail and phone logs, and matter numbers or client contact details on file. The language work can run on the firm's own computer.

You stay in control

The timekeeper edits and releases their own entries.

Connects to

  • Calendar
  • Email
  • Phone log
  • Billing software

Questions we would ask you

  • How do timekeepers record time now (timer, end of day, end of month), and how much do you think is lost?
  • Which calendar, mail and phone systems do you use, and do lawyers take client calls on personal mobiles?
  • Do matter numbers appear in email subjects or folder names?
  • What minimum increment and estimation rules does the firm use for email and document work?
  • Which billing system do you use, and on which plan?
Discuss this further with the Live Agent

Documents & compliance

Contract review against your playbook

An incoming NDA or lease is compared clause by clause with the positions your firm normally takes, and comes back with a marked-up draft and a short issues list.

Starts when

A contract is saved to the review folder.

What happens

  1. Identify each clause and compare it with your standard position
  2. Mark clauses that are missing, unusual or outside your fallback terms
  3. Suggest replacement wording from your own clause library
  4. Summarize the open points in order of importance

In more detail

An automation like this reviews an incoming contract, such as an NDA or a lease, against the positions a firm normally takes. When a contract is saved to a review folder, it splits it into clauses, identifies each clause type, and compares each one with the firm's written playbook of preferred wording, fallbacks and walk-away points. It marks clauses that are missing, unusual or outside the fallbacks, suggests replacement wording from the firm's own clause library, and returns a marked-up copy plus a short issues list in order of importance. The reviewing attorney accepts, rewrites or rejects every suggestion. The firm needs a written playbook and approved clause wording for the contract types it wants covered; writing those down is most of the work. The language work can run on the firm's own computer so contract terms stay in the office.

You stay in control

The reviewing attorney accepts, rewrites or rejects every suggestion.

Connects to

  • Document storage
  • Word processor
  • Clause library

Questions we would ask you

  • Which contract types do you review most, and how many a month?
  • Do you have a written playbook or checklist for any of them? Who would write one?
  • Where does your approved clause wording live today?
  • Which side do you usually represent (landlord or tenant, vendor or customer)?
  • Do you need true Word tracked changes, or are comments plus an issues list enough?
Discuss this further with the Live Agent

Documents & compliance

Medical-record chronology for injury matters

A stack of provider records becomes a dated timeline of visits, findings and treatment, each line linked to the page it came from.

Starts when

Records are received from a provider.

What happens

  1. Make the records searchable and sort them by provider and date
  2. Extract each visit, complaint, finding, procedure and prescription
  3. Build one chronology across all providers with page references
  4. List gaps in treatment and records that appear to be missing

In more detail

An automation like this turns a stack of medical records in an injury matter into a dated chronology. It makes scanned records searchable, sorts them by provider and date, and pulls out each visit, complaint, finding, procedure and prescription, with every line linked to the page it came from. It also lists gaps in treatment and records that seem to be missing. A paralegal checks every entry against the source page before the chronology is used. Medical records are protected health information, so this would run only on the firm's own hardware and needs written privacy controls. Texas law can treat anyone holding such records as covered by state medical-privacy rules. Several established vendors sell this service, so Fenrir Labs would usually suggest comparing them first, or starting with a simpler page-linked index without AI extraction.

You stay in control

A paralegal verifies entries against the source pages before the chronology is relied on.

Connects to

  • Scanner
  • Document storage

Questions we would ask you

  • How many injury matters do you open a year, and how many pages of records does a typical file have?
  • Who builds chronologies today, and how long does one take?
  • How do records arrive: fax, mail, portal downloads, a records-retrieval vendor?
  • Do you already use EvenUp, Supio or a similar service, and what do you pay per case?
  • Do you represent plaintiffs, defendants or both? Do you ever work for a hospital or insurer?
Discuss this further with the Live Agent

Customer service

Client status updates that actually go out

Every active client gets a regular, plain-language update drafted from the docket, the file notes and what is scheduled next.

Starts when

A schedule you set per matter, or a new docket event.

What happens

  1. Gather what changed since the last update
  2. Draft a short update in plain language with the next step and date
  3. Note anything the firm is waiting on from the client
  4. Queue the draft for the responsible attorney

In more detail

An automation like this drafts regular, plain-language status updates for each active client matter. On a schedule the firm sets per matter, or after a significant court event, it reads what changed in the practice-management system since the last update, such as hearings, filings, completed tasks and file notes, and drafts a short message: what happened, what happens next and when, and anything the firm is waiting on from the client. Internal notes are filtered out before drafting. Each draft goes to the responsible attorney, who edits and approves it and sends it from their own email. The system never sends anything by itself. It needs read access to the practice-management system and current client contact details. Firms on some practice-management plans already have AI that drafts client updates, so the first question is what the firm already has.

You stay in control

The attorney reviews each update before it is sent to the client.

Connects to

  • Practice management
  • Docket
  • Email

Questions we would ask you

  • How many active matters does each attorney carry, and how often do clients call for status?
  • Do you send any regular updates today? What stops them going out?
  • Which practice-management product and plan do you use, and are notes and tasks kept current?
  • How do you mark internal notes or work product in the file?
  • Do clients prefer email, portal or text?
Discuss this further with the Live Agent

Finance & billing

Pre-bill review before the partner sees it

Before invoices go to the partner, entries are checked against the client's billing guidelines and vague narratives are rewritten for approval.

Starts when

The monthly billing run starts.

What happens

  1. Check each entry against the client's billing rules and rate agreement
  2. Flag block billing, duplicates and entries on closed matters
  3. Suggest a clearer narrative where the description is too thin
  4. Total what is ready to bill and what needs attention, by partner

In more detail

An automation like this does the first pass of a firm's monthly prebill review. When the billing run starts, it reads the draft invoices and their time and expense entries, checks each entry against the client's billing rules and agreed rates from the engagement letter and any client guidelines, and flags block billing, duplicates, wrong rates and entries on closed matters. Where a narrative is too thin, it suggests clearer wording using only facts already in the entry, never adding work. It totals what is ready to bill and what needs attention for each partner. The billing partner decides every write-down and approves each invoice; the system never changes or sends a bill. It needs read access to draft bills, through the billing software's API or an export, and copies of engagement letters. The work can run on the firm's own computer.

You stay in control

The billing partner decides every write-down and approves each invoice.

Connects to

  • Billing software
  • Engagement letters
  • Email

Questions we would ask you

  • How many invoices go out a month, and how long does prebill review take the billing partner?
  • Which billing system and plan do you use, and can it export draft bills with line items?
  • Do any clients have written billing guidelines or require LEDES or UTBMS codes?
  • What are your most common write-downs, and why?
  • What does a good narrative look like at your firm? Can you share examples?
Discuss this further with the Live Agent
Ask the assistantWhat could we automate for you?