Loading
Loading
One pipeline for every new inquiry: structured intake records, conflict checks prepared before an attorney opens the file, documents filed automatically, deadlines calendared only after a lawyer confirms them, and clients kept informed at each stage. Privileged content stays inside firm systems.
Solution blueprint: a system Callimachus Forge can architect for your operation. Not a case study, and not a claim of past deployment.
At a glance
Stack candidates
n8n · Typeform · Gmail · Postgres · Google Calendar · GoHighLevel · Twilio · Slack
System run
A simulated run of the exact steps documented on this page, from first trigger to final handoff. Nothing here is live client activity.
Simulation · Loops
The problem
New matters arrive by phone, email, referral, and web form, and each one starts a scramble. Someone retypes the caller's story into a memo. Conflict checking means emailing names around the office and searching old files, so matters sit for days before anyone can say yes. Client documents arrive as attachments and paper, and filing them is a side task that slips. Deadlines get entered by hand into individual calendars, which means they live with one person. Clients call for updates because nobody has time to send them. None of this is legal work, but attorneys and staff spend real hours on it, and a missed conflict or a missed date is the kind of error a small firm cannot absorb.
The typical firm here has two to eight attorneys, a paralegal or two, and an office manager who owns intake by default. Inquiries come through the main phone line, a contact form, and individual attorney inboxes. Client and matter history lives in a practice management system, an old spreadsheet, or closed-file boxes, which is what makes conflict checking slow. Calendars are personal Outlook or Google accounts. Confidentiality is not optional: intake details are often privileged or sensitive from the first call, so any system has to keep matter content inside firm-controlled storage and out of general-purpose marketing tools.
The manual process
The automation opportunity
Every inquiry, from any channel, feeds one pipeline. The system reads the inquiry, pulls out the people and companies involved, the type of matter, and any dates mentioned, and creates a clean intake record. It searches the firm's client and matter history for those names, including near matches and misspellings, and hands the attorney a prepared conflict memo instead of a blank question. Nothing opens until a lawyer signs off. After approval it drafts the engagement letter, files incoming documents, proposes deadlines for a lawyer to confirm, calendars confirmed dates with reminders, and sends clients short status notes at each stage. Privileged content stays in firm systems throughout.
Architecture
Every step is typed: triggers, AI decision points, business logic, actions, data, and the human checkpoints that keep judgment where it belongs.
AI components
Human checkpoints
A web intake form (Typeform), a dedicated intake mailbox (Gmail), or a manual entry by staff fires a webhook into n8n. Every channel lands in the same pipeline with a timestamped intake ID.
A language model parses the form text, email body, and attachments to pull client name, adverse and related parties, matter type, jurisdiction, and any dates mentioned. Missing or ambiguous fields are flagged for staff.
The structured record is written to firm-controlled Postgres and documents go to access-restricted matter storage. Only contact details and a stage label sync to the CRM; matter content never leaves firm systems.
n8n runs exact and fuzzy matches of every extracted party against the firm's client and matter history in Postgres, including spelling variants and aliases, and compiles hits into a memo with matter numbers and roles.
The attorney gets the conflict memo and intake summary in one message. They approve, decline, or request more information. The pipeline blocks until this decision is recorded; declines trigger a courteous close-out.
On approval, n8n creates the matter record, builds the folder structure, files intake documents into it, and drafts an engagement letter from the firm template into the attorney's drafts for personal review and sending.
A model rereads the intake and documents for triggering events and dates, then proposes calendar entries, quoting the source text for each. It proposes only; it never computes limitation periods on its own authority.
An attorney or paralegal confirms, edits, or rejects each proposed date. Unconfirmed dates stay visibly flagged in the digest until resolved, so a proposal can never quietly become a deadline.
Confirmed dates are written to a shared Google Calendar with tiered reminders and mirrored in Postgres, so deadlines survive staff changes and stay visible to the whole matter team, not one inbox.
When a matter changes stage, a templated email or SMS goes out through GoHighLevel or Twilio. Templates carry status language only, such as documents received or review scheduled, and never include matter substance.
A recurring summary posts to the firm's Slack channel or inbox: new intakes, conflict reviews waiting, unconfirmed deadline proposals, and dates in the next two weeks, so nothing depends on someone remembering to check.
Engineering notes
Benefits are qualitative by design. We do not publish numbers we cannot verify.
The free audit maps your actual tools and process against this architecture, and tells you honestly what is worth building.