We work remotely with solo attorneys and small law firms across the United States. It fits firms of one to ten attorneys without a dedicated IT person. The work we do is concrete and bounded: setting up AI tools that respect client confidentiality, building prompt libraries for the kinds of work you do every week, and training your team to use them without surprising you on a bar complaint.
We are not lawyers and we don't give legal advice. We're the people who configure the tools, document how to use them, and stay on call when something changes. We take no commissions from any AI vendor, so when we recommend Claude over ChatGPT (or the reverse), it's because of the work, not the kickback.
Watch the overview
What this looks like for law firms.
Recorded for our Portland-area clients. Everything in it applies wherever you are.
Read the transcript
Look, if you're running a small law firm right here in the Portland metro area, you have undoubtedly seen the endless AI hype. You suspect you probably need to do something about it, but you just don't trust the usual tech consultants. I completely get it. Today, let's just skip the hype altogether and look at the grounded, practical reality of what this actually means for a firm your size. Let's start by looking at where the hours actually go in a 1-10 attorney firm. It's the routine stuff that really gets you. First drafting absolutely drains senior attorney time. Intake and conflict checks are incredibly repetitive. Discovery on small matters is often too expensive to outsource, but takes way too much time in-house. And all the while, client communications just continue to pile up. So where does AI actually fit in to give you those hours back? Well, it mainly lives in a few very specific workflows. Step one is first draft document work. Think about generating demand letters or routine motions directly from your existing templates and matter facts. Step two is intake and conflict check automation. You can set up workflows that capture details consistently, immediately surface any conflicts against your client list, and then seamlessly route the file to the right person. Now let's just pause right here for a second. Because to truly evaluate this tool for your practice, we need to have a very frank conversation about its absolute limits. Here is what this is plainly not for. First it is not for replacing attorney judgment. AI might handle that first draft, but an attorney reviews and signs. Always. Second, it's not a source of authority. You absolutely still have to verify everything in Westlaw or Lexus. And finally, this is never for sending confidential client data to consumer-grade AI accounts. Client data has to be protected with enterprise controls. Period. Look, if you decide to explore this, you need a partner whose incentives actually align with yours. So here are our firm commitments to you. First, we take no software commissions or vendor kickbacks. Ever. Second, we work strictly on time and materials, with a written, not-to-exceed cap quoted before the work even starts. And third, we will tell you flat out when you simply don't need AI. If you're ready to stop managing paperwork and actually get those hours back in your week, let's just have a conversation. Book a free consult using the short form right here on this website.
Where the hours go.
These are common recurring time sinks for law firms, wherever they are.
- Document drafting takes a disproportionate amount of senior attorney time.
- Intake and conflict checks are repetitive and prone to drop-off.
- Discovery review on small matters is too expensive to outsource and too time-consuming to do well in-house.
- Client communication piles up between matter milestones.
- Marketing content and SEO sit at the bottom of the priority list permanently.
Where AI actually fits.
The specific projects we offer law firms. Each is scoped, quoted in writing, and capped at a not-to-exceed price.
First-draft document work
Demand letters, engagement letters, discovery responses, and routine motions drafted from your templates and your matter facts. Reviewed by an attorney before anything goes out. The time savings are real and the workflow is auditable.
Intake and conflict-check automation
AI-assisted intake forms that capture matter details consistently, surface conflicts against your existing client list, and route to the right intake person. Connects to your case management system where one exists.
Confidentiality-safe AI setup
Enterprise-tier accounts (Claude for Work, ChatGPT Enterprise, Microsoft Copilot) configured so client data does not train public models. Written documentation of what's allowed and what isn't, suitable for your bar's competence and confidentiality rules.
Prompt library for recurring work
Tested, documented prompts for the document types and analyses you do every week. Stored where your team can find them and updated as the underlying models improve.
Team training
Live, hands-on sessions on Google Meet (in person for Portland-area clients), on your actual tools and your actual matter types. Your staff learns by doing, with us working alongside. Google Meet sessions are recorded, and the recordings stay with the firm for onboarding new attorneys and paralegals.
ChatGPT for lawyers: what it's good for.
Most small firms already have someone using ChatGPT, often on a personal account. That's the part to fix first. Used the right way, it's a fast first-draft tool. Used the wrong way, it's a confidentiality problem.
Good uses
- First drafts of client letters, emails, and routine motions from your own templates, which an attorney then edits.
- Summarizing a long document or deposition transcript so you know where to read closely.
- Turning your notes into a plain-English update for a client.
- Checklists and intake questions for the matter types you handle every week.
Keep out of it
- Client names, facts, or documents in a free or personal account.
- Case citations you haven't checked. ChatGPT can invent cases that look real. Verify everything in Westlaw, Lexis, or Fastcase.
- Anything you'd sign without reading it first.
Set it up the safe way
ChatGPT Business (the plan formerly called Team) excludes your data from model training by default. Free and personal accounts can use conversations for training unless someone turns that off in settings. The ABA's Formal Opinion 512 (July 2024) says lawyers may use these tools if they understand them, check the output, protect client information, and bill only for time actually spent. It also calls for a client's informed consent before their information goes into a tool that learns from what you type. Check your own state bar's guidance. We set up the tools and write the firm policy; the legal calls stay yours.
What this isn't for.
Where AI isn't the right answer in this industry. Knowing this up front saves everyone time.
- Replacing attorney judgment. AI drafts. Attorneys review and sign. Always.
- Legal research as a source of authority. AI can suggest where to look. Attorneys verify in Westlaw, Lexis, or Fastcase before relying on anything.
- Sending confidential client data to consumer-grade AI accounts. Not without the right enterprise controls and a clear written policy.
Where we work with law firms.
Remotely, across the United States. We're based in Portland, Oregon, so clients nearby can have the first visit on-site. How that works.
FAQ
Questions from law firms.
- Is using AI for legal work compliant with bar rules?
- It can be, with the right setup and the right human oversight. Bar associations, including the ABA and state bars such as Oregon's, have issued guidance affirming that lawyers may use AI tools provided they meet their duties of competence, confidentiality, and supervision. We configure tools so client data is not used to train public models, document what's allowed in writing, and structure workflows so an attorney reviews any AI-generated output before it goes out. We are not your ethics counsel, but we know the guidance and we won't set up something that obviously conflicts with it.
- Which AI tool is best for small law firms?
- It depends on the work. Claude (from Anthropic) tends to do better on careful drafting and document review and has strong enterprise data controls through Claude for Work. ChatGPT Enterprise is broader and integrates with more third-party tools. Microsoft Copilot makes sense for firms already deep in Microsoft 365. There are also legal-specific tools (Harvey, CoCounsel, Spellbook) that may fit larger firms but are typically overkill and overpriced for solos and small offices.
- How much do small law firms spend on AI consulting?
- A law firm can start in one of three ways: a $600 written audit, a single implementation project at $150/hour with a written not-to-exceed cap (most run $500 to $5,000), or a series of small projects spread over a few months. Ongoing advisory is $600 a month with a 3-month minimum, and a small firm may not need it once the initial setup is in place.
- Can AI help with discovery review on small matters?
- For document review specifically, the practical answer is yes for clearly relevant document classification and initial privilege screening, but only as a first pass with attorney verification. For matters where the document volume is in the hundreds or low thousands, we can set up a workflow using Claude or ChatGPT Enterprise with explicit safeguards. For larger document sets, the right answer is usually a purpose-built eDiscovery tool, and we'll say so.
- Can lawyers use ChatGPT?
- Yes, with care. The ABA's Formal Opinion 512 (July 2024) allows lawyers to use generative AI if they understand the tool, verify what it produces, protect client confidentiality, and bill fairly. In practice that means a business account that doesn't train on your data, a written firm policy, and an attorney reviewing every draft. Check your state bar's guidance as well.
Other industries we work with.
- AI for real estate offices
- AI for accounting and bookkeeping practices
- AI for gyms and fitness studios
See all industries or see where we work.