AI Playbook for Attorneys: 7 Places to Use AI (And the One Place That Will End Your Career)

1,200 attorneys.

One $109,700 sanction.

That is the count Damien Charlotin keeps at HEC Paris. He runs a global database tracking every court sanction issued for AI fabricated citations. About 800 of the 1,200 are from US courts.

“Recently we had 10 cases from 10 different courts on a single day,” Charlotin told NPR in April.

The $109,700 belongs to two Oregon lawyers, Stephen Brigandi and Tim Murphy, who filed AI hallucinated briefs in a family winery dispute. Judge Mark D. Clarke did not just sanction them. He dismissed the case with prejudice. The client lost. The lawyers paid. Both names are on the docket forever.

Carla Wale runs the law library at the University of Washington School of Law. Her title is Associate Dean of Information and Technology and Director of the Law Library. She trains the next generation of attorneys on AI ethics. Her rule for everyone reading this is one sentence:

“Whatever the generative AI tool gives you, as in, ‘Look at these cases,’ you, under the rules of professional conduct, you have to read those cases.”

That is the cliff edge.

If you read those numbers and felt either smug or scared, this article is for you. The sanctions story is real. It is also half the story. The other half, the half nobody is writing about, is what is happening at the firms that figured out where AI actually pays. They are not the firms making headlines. They are the ones quietly recovering 6 to 8 billable hours per attorney per week, then layering AI into every other stage of their operations until the whole firm runs leaner than the firm down the street.

This is the playbook. We start with the one habit that ends careers. Then we walk through seven specific tools that recover hours. Then we map every stage of a law firm’s lifecycle and show you where AI fits at each one, including the stages most attorneys never think about applying it to.

The One Habit That Will End Your Career

Read this carefully. It is not what you think.

The cliff is not “using AI for case law.” It is not “using AI for legal research.” It is not even “using AI.”

The cliff is filing AI output without reading what it gave you.

Steven Schwartz did not get sanctioned for using ChatGPT in Mata v. Avianca (2023). He got sanctioned for filing six fabricated cases without opening any of them. The cases ChatGPT invented sounded reasonable. The citations looked properly formatted. He did not click through. He did not pull the cases on Westlaw. He filed.

That is the cliff. Schwartz got sanctioned $5,000. Brigandi and Murphy got hit for $109,700. The 1,200 cases in Charlotin’s database all share the same fact pattern. AI gave the lawyer something. The lawyer filed it. The cases were not real. The court found out.

Federal Rule of Civil Procedure 11 still requires you to certify that every legal contention is warranted by existing law. Model Rule 1.1 still requires competence. Nothing about generative AI relaxes either rule. AI is just a faster way to be wrong if you do not verify.

Wale’s quote is the operating principle. Whatever the generative AI tool gives you, you have to read those cases. The AI is not the lawyer. You are. The AI giving you a list of cases is the same as a paralegal handing you a list of cases. You verify. Every time. No exceptions.

Hold this distinction. AI for citation work is fine. There are legal research tools built specifically to never hallucinate, and we recommend one of them later in this article. The line is not at the tool. The line is at the keyboard, the moment you decide whether to file something you have not read.

The seven tools below all sit comfortably on the safe side of that line. Each one does work that does not require you to read every word of a Westlaw return or every page of a deposition. Each one cuts your hours and your malpractice exposure stays at zero, because the work AI is doing is not the work that gets you sanctioned.

Before we get there, one operating note. Most attorneys we talk to want help mapping this for their specific practice. We do that on a free 30 minute call. Top three AI use cases ranked for your firm by ROI, estimated billable hours reclaimed per attorney, a 90 day implementation roadmap on a single page you can show your partners. No pitch. Book it at https://enzeti.com/calendar/ if you want one. We onboard four firms a quarter.

Now the seven.

The Re-frame: AI as Operations Layer, Not Brain

The mental model that changes everything is this.

Your firm already has layers. Partners do legal judgment and client relationships. Associates draft and research. Paralegals organize and summarize. Court reporters transcribe. Each layer does work that the layer above it does not have time for.

AI is a new layer. It sits below the paralegal layer. It does the work that even paralegals find tedious. Transcript ingestion. Boilerplate discovery responses. Medical record extraction. Contract clause comparison. The work that, if you watched a paralegal do it for eight hours, you would say “this should be a machine.”

It is not the lawyer. It is not the associate. It is a new layer of operations that sits underneath everything else and never sleeps.

The firms winning with AI are not the ones using it to “be smart.” They are the ones using it to remove the operational friction that has always existed between billable thought and billable output. The legal thinking still happens at the partner and associate level. The drafts still need attorney review. The strategy still requires human judgment. AI just removes the eight hours of grinding between point A and point B.

Hold that mental model as you read the seven. Each one is an operations layer placement. None of them is a brain transplant.

Use Case 1: Deposition Transcripts and Summaries (Parrot AI)

The use case: get the transcript and the summary of any deposition fast enough to act on it the same week.

The pain it removes: the standard timeline for a deposition transcript is days to weeks. The summary either does not exist or eats a paralegal’s afternoon. You walk out of a deposition, you wait, you read the cold transcript six days later, you have already lost the thread.

The tool: Parrot AI. They built proprietary language models trained specifically on legal and insurance vocabulary. Transcripts are ready in 90 minutes. AI generated summaries are ready in seconds, not hours.

How to use it without going overboard: use Parrot for transcription and first pass summary. You still read the transcript before you cite it in a motion. You still verify the summary against the actual transcript before relying on it. Wale’s rule applies. The AI surfaces. The attorney verifies.

Practice area example: family law, custody dispute. Opposing party’s deposition runs four hours. Custody hearing is eight days out. Old workflow: wait for the transcript, read it cover to cover, hand mark contradictions, brief the client. New workflow: Parrot transcript in 90 minutes, AI summary identifies three contradictions with the client’s affidavit, you spend the saved time prepping cross instead of skimming.

Time savings on a family law case with one custody deposition: roughly 6 to 8 hours of paralegal and attorney time recovered.

For more on how intake quality drives custody case outcomes, see our family law intake article.

Use Case 2: Discovery Response Drafting (Briefpoint)

The use case: drafting responses to interrogatories and requests for production.

The pain it removes: discovery response drafting is one of the most loathed parts of civil litigation. Most associates dread it. Most paralegals burn out on it. Briefpoint estimates that discovery response drafting costs the average firm $23,240 per attorney per year, based on an associate at $150,000, 20 cases per year, four discovery sets per case, and 3.5 hours per response set.

The tool: Briefpoint. They auto draft discovery responses. You upload the propounded discovery, the AI translates the questions into client friendly language for your client to answer, then plugs the responses back into the formal document. Output is a Word document ready to review and serve.

Their published numbers: 87% time reduction on discovery response drafting. 2 to 4 hours saved per response set. 30+ hours saved per case. 1,500+ law firms using it.

How to use it without going overboard: Briefpoint drafts. The attorney reviews objections and edits any nuanced answer before serving. Boilerplate responses get the AI treatment. Anything substantive gets attorney eyes. Wale’s rule applies inside the firm just like it applies outside.

Practice area example: civil litigation, commercial dispute, four discovery sets in one case. Old workflow: associate spends 14 hours drafting across four sets. New workflow: associate spends 3 hours reviewing AI drafts, partner spends one hour on objections. Net recovery: 10 hours per case, 30+ hours across the case lifecycle.

The economics are not subtle. At a $300 blended hourly rate, 30 hours of recovered associate time per case is $9,000 of recovered capacity. Across 20 cases per associate per year, that is $180,000 of capacity recovered without hiring.

For more on the staffing implications of recovered capacity, see our intake coordinator staffing article.

Use Case 3: PI Demand Letters from Medical Records (Supio)

The use case: building medical chronologies, economics tabs, and demand letters for personal injury cases.

The pain it removes: pre litigation work in PI is the most time intensive part of the practice. Medical record review alone can run 20 to 40 hours per case. Building a chronology is another 10. Drafting the demand on top of that is another 10 to 20. The whole pre litigation package can eat 50 to 80 hours of paralegal and attorney time before the demand goes out.

The tool: Supio. Built specifically for PI law firms. AI medical record review. Medical chronology generation. Economics tab. Demand letter drafting. They published a case study from J. Chrisp Law showing 80+ hours saved per case on chronology, economics, and demand work.

Their published numbers: 96.6% extraction accuracy on medical records. $500 to $1,000 saved per case in raw labor. SOC2, HIPAA, and GDPR compliant for sensitive medical data.

How to use it without going overboard: Supio drafts the chronology and the demand. The attorney reviews the chronology against the actual records before relying on it for negotiation. The demand letter gets attorney level review for tone, valuation, and case theory before it goes out. The AI does the extraction. The attorney does the strategy.

Practice area example: PI specifically. Auto crash case, soft tissue, three years of treatment records across five providers. Old workflow: paralegal spends 30 hours building the chronology, attorney spends 15 hours drafting the demand. New workflow: Supio outputs both in hours, paralegal spends 4 hours verifying the chronology against the actual records, attorney spends 4 hours editing the demand for case theory.

Net recovery: 30 to 40 hours per case. Across a PI firm running 100 cases a year, that is 3,000 to 4,000 hours of recovered capacity.

For more on PI intake quality and case selection, see our rideshare accident intake article and motorcycle accident intake article.

Use Case 4: Contract Review in Microsoft Word (Spellbook)

The use case: first pass contract review for transactional, family law, and small business work.

The pain it removes: contract review is line by line work. Reading a 30 page contract for the first time, identifying the bad clauses, flagging the missing protections, comparing to your firm’s standard. Most attorneys do this in Word and it is slow.

The tool: Spellbook. It runs as an add in inside Microsoft Word. You open your contract, you click Spellbook in the ribbon, the AI does a first pass review. Risks are flagged. Suggested redlines are inserted. You stay inside Word the whole time. No uploading, no copy paste, no switching tabs.

Their published numbers: up to 7 hours saved per contract on average. 70% reduction in review time. Used by 4,000+ firms across 80+ countries. SOC 2 Type II compliant. Does not train its models on your client documents. Pricing starts at $99 per user per month.

How to use it without going overboard: Spellbook flags. The attorney decides. The first pass review is AI. Every flag and every suggested redline gets attorney review before going to opposing counsel or to the client. The AI suggests. The attorney negotiates.

Practice area example: family law, prenuptial agreement. Old workflow: attorney spends 6 hours reviewing the proposed prenup line by line against state law and the firm’s own template. New workflow: Spellbook flags the 9 risk clauses and proposes redlines in 8 minutes, the attorney spends 2 hours reviewing flags and tightening the language for the specific clients.

Time savings on a single prenup: 4 hours. Across a family law firm doing 20 prenups a year: 80 hours.

Or the same flow for a small business attorney reviewing 50 commercial contracts a year. Conservative math at 4 hours saved per contract: 200 hours of recovered capacity per year.

For more on serving family law and small business clients efficiently, see our family law intake article and employment law intake article.

Use Case 5: Legal Research with Citation Verification (Paxton AI)

The use case: legal research that does not hallucinate citations.

The pain it removes: this is the cliff problem in reverse. Generative AI is fast at research. Generative AI also fabricates citations. Most attorneys solve this by not using AI for research at all, which means they pay full Westlaw or Lexis rates and burn associate hours on every memo.

The tool: Paxton AI. They built their citator specifically to verify every citation against actual primary sources. The citator is patent pending and was featured in Law.com. Paxton commits to no hallucinated citations. Every claim is backed by a source you can click into and read.

Compliance: SOC 2, ISO 27001, and HIPAA. Data is encrypted. Never used for model training. Built for attorney workflows.

How to use it without going overboard: even with Paxton, Wale’s rule still applies. The citator surfaces the cases. The attorney reads the cases. Paxton is a faster Westlaw with a better drafting layer on top. It is not a substitute for opening the case and reading the holding. The AI surfaces. The attorney verifies.

Practice area example: criminal defense, motion to suppress on a Fourth Amendment search question. Old workflow: associate spends 6 hours on Westlaw building the research memo. New workflow: associate runs the natural language query in Paxton, the citator returns the relevant cases with summaries and links to the underlying opinions, the associate reads the four most relevant cases in full and writes the memo. Time on Westlaw: zero. Time recovered: 3 to 4 hours per memo.

This is the use case where the cliff and the playbook meet. You can use AI for research. You just have to use a tool built around verification, and you still read the cases.

Use Case 6: Judge Analytics and Motion Outcome Prediction (Trellis)

The use case: data on how a specific judge has actually ruled on the motion you are about to file.

The pain it removes: the unknown. Every motion practice attorney has experienced filing a motion in front of a judge whose tendencies they did not know, and losing on a procedural angle they could have anticipated. Most attorneys deal with this by asking around the bar association, which is unreliable, slow, and political.

The tool: Trellis. They built the largest US trial court database, with structured data on judges, attorneys, filings, and outcomes. Their Judge Analytics dashboard shows you how a specific judge has ruled on specific motion types over the past several years. Their Law Firm Intelligence product launched seven new analytics dashboards in 2026 alone.

What you get: a judge’s grant rate on motions for summary judgment. Discovery motion grant rates. Comparison against other judges in the county and the state. The ability to see, before you file, whether your motion has a 70% historical chance of being granted in front of this specific judge or a 12% chance.

How to use it without going overboard: Trellis is a probability layer on top of attorney judgment. It is not a substitute for understanding the case. The data tells you what the judge has done. The attorney decides what to do with that information.

Practice area example: civil litigation, motion for summary judgment. Old workflow: attorney decides whether to file based on case theory and instinct. New workflow: attorney pulls the Judge Analytics dashboard, sees that this judge has granted MSJs in 58% of similar motions, sees the typical reasons for denial, tightens the brief on those points before filing.

Or the inverse, where the data shows a 12% grant rate. Old workflow: file anyway, lose, look bad to the client. New workflow: see the data, decide not to file, save the client $15,000 in attorney fees and keep your credibility intact for a stronger motion later.

The time savings here are not just hours. They are case strategy.

Use Case 7: Real Time Intake Call Coaching (eNZeTi)

The use case: real time AI prompts for the human running your intake calls. Live call scoring. Post call dashboards. Weekly improvement loops.

This is what we built. Full disclosure: this section is about us. We saved it for last because every other tool above moves work that has already converted into a client. eNZeTi is the tool that decides whether the work ever exists in the first place.

The lead conversion principle is universal. We will keep this section short and we will tell you the same thing we tell our clients on the call.

Here is the proof we use. Cameron runs a sales team at a fast growth education company called BecomeViral. Not a law firm. We dropped our augmentation model into his team’s call workflow. Same people. Same scripts. Same compensation.

30 days later: 2x production.

We are showing you a non legal example on purpose. Augmentation is not a legal vertical magic trick. It is a “humans converting leads on calls” principle. Your intake team does the same job Cameron’s team does. They qualify, they build trust, they set the next step. The AI tells the human, in real time, what the best version of the human would say next.

How to use it without going overboard: the AI never replaces the intake coordinator. It coaches them. The coordinator decides which prompt to use and which to ignore. The AI surfaces. The human delivers.

Where this fits in the playbook: most attorneys we talk to assume their intake is fine. Then we look at the actual call data and the conversion rate is 30 to 40% lower than they think. Augmentation is the fastest path to closing that gap without hiring.

If you want to see what your firm’s number actually is, we audit it free. Book at https://enzeti.com/calendar/.

For practice area specific intake plays, see our articles on medical malpractice intake, immigration intake, and DUI intake.

The Full Law Firm AI Lifecycle: Where AI Fits at Every Stage

Most articles about AI in law firms stop at the seven tools. The firms pulling ahead are not stopping there. They are layering AI into every stage of how the firm runs, from the moment a prospect first hears about you to the moment a closed client refers their cousin three years later.

This is the map. Eight stages of the law firm lifecycle. Where AI fits at each stage. Specific tools, specific time savings, specific impact on revenue.

Read this carefully. Most firms are using AI in two of these eight stages. The firms running circles around them are using AI in six.

Stage 1: Find the Right Clients (Marketing and Demand Generation)

Before anyone calls your firm, they have to know you exist. The marketing stage is where most law firms still burn the most money for the least return. AI fixes a specific subset of that.

SEO content production. The single highest leverage AI use case for law firm marketing. Use Claude or ChatGPT to draft long form articles on your practice area. Attorney reviews and edits for accuracy and voice. Publish at 4 to 8 articles per month instead of 1. Time per article drops from 8 hours to 2.

Local search optimization. Google Business Profile updates, review responses, location pages. Tools like BrightLocal automate the workflow. AI drafts review responses for the attorney to approve. Saves 2 to 4 hours per week.

Paid ad creative. Use AI for headline variations, ad copy, and image generation. NEVER use AI for compliance language or disclaimers. Test 10 variations of a Google Ads headline in the time it used to take to write 2.

Time saved at this stage: 8 to 15 hours per week of marketing labor for a typical small firm. That is one part time hire of capacity recovered.

Where this connects to eNZeTi: the call data tells you which marketing source actually closes. We surface the conversion rate by lead source so you stop spending on the channel that brings in tire kickers.

Stage 2: Qualify Before the Call (Lead Enrichment and Pre-Call Research)

Most firms do zero pre call research. The intake coordinator picks up cold and asks the same five questions every time. The result is a 25 minute call that could have been a 12 minute call if anyone had spent 3 minutes preparing.

Lead enrichment. When a contact form fires, AI tools enrich the record with public data. LinkedIn role, employer, location, basic profile. Now your intake coordinator opens the call already knowing this is a senior partner at a regional firm, not a college student.

Pre call brief generation. AI drafts a 1 paragraph pre call brief. Who they are, what their inquiry was about, what to confirm in the first 90 seconds, which practice area lead they map to. Coordinator reads the brief in 60 seconds before dialing.

Intent scoring. Hot vs warm vs cold based on form responses, time of submission, source, and what they typed in the message field. Hot leads get called within 5 minutes. Warm leads within 30. Cold leads get an automated nurture sequence.

Time saved at this stage: 15 to 20 minutes per prospect by shortening the discovery portion of the call. Across a firm taking 30 inquiries a week, that is 7 to 10 hours of intake time recovered weekly.

Where this connects to eNZeTi: the lead enrichment feeds directly into the live call coaching. Coordinator picks up with context, the AI prompts adjust based on what was already known, and the call moves twice as fast through qualifying questions.

Stage 3: Convert on the Call (The Intake Conversation Itself)

This is the conversion moment. Everything before this stage is preparing for it. Everything after this stage is dependent on whether it converts.

Real time AI prompts. While the coordinator is on the call, AI listens, scores, and surfaces the next best thing to say. Objection comes up about price. Prompt suggests the framing that has converted that objection 73% of the time at peer firms. Caller mentions they are also talking to two other firms. Prompt surfaces the differentiator language that has worked before.

Live call scoring. Each call gets a score in real time on six dimensions. Empathy. Question quality. Pace. Compliance. Closing. Recap. Coordinators see their score the moment the call ends.

Post call dashboards. Daily dashboard showing call volume, conversion rate by coordinator, conversion rate by lead source, conversion rate by practice area, average call length. Everything actionable in one place. No more “I think our intake is doing fine” guesses.

Time saved at this stage: not the right metric. The right metric is conversion lift. Cameron’s team doubled production in 30 days. Law firms running our model see typical conversion lifts of 30 to 40% within 60 days, often without adding a single new lead source.

Where this is eNZeTi: this is the entire reason we exist. If you only adopt AI in one stage of your firm’s lifecycle, this is the highest leverage stage to start. Every other stage downstream depends on whether the call converted.

Stage 4: Close the Engagement (From Yes to Signed)

Some firms lose 20 to 30% of their “yes” calls between the moment the prospect agrees and the moment they actually sign the fee agreement. The friction is administrative. AI removes it.

Auto generated fee agreements. The coordinator clicks “send fee agreement” inside the practice management system. AI populates a templated fee agreement with the client’s specific facts, the practice area scope, and the rate structure. Spellbook can do this directly in Word. So can Filevine and other practice management systems with AI features.

E-signature with smart reminders. Fee agreement goes via DocuSign or a similar tool. Smart reminders use AI to time follow ups based on when the client is most likely to be on their phone. Sign rate goes from 65% to 85% with no human follow up.

Conflict checks. AI scans the prospect’s name, employer, and adverse parties against your client database in seconds instead of the 20 minutes it takes a paralegal to do it manually.

Time saved at this stage: 1 to 2 hours per new client. For a firm signing 10 new clients a month, that is 10 to 20 hours of paralegal time recovered monthly. The bigger win is the conversion rate from “yes” to “signed,” which can move from 70% to 90% with the right automation.

Where this connects to eNZeTi: we hand the conversion off to the close, then surface the dropout rate at this stage so you know exactly where the leak is. Most firms do not even measure this stage.

Stage 5: Do the Legal Work (The Matter Itself)

This is where the seven tools above live. Once the client is signed, the work begins, and every practice area has its own AI leverage points.

For litigation: Parrot for depositions, Briefpoint for discovery responses, Trellis for judge analytics, Paxton for legal research.

For PI specifically: Supio for medical chronologies and demand letters.

For transactional: Spellbook for contract review.

For any practice area requiring research: Paxton for citation verified legal research.

The full breakdown is in the seven sections above. The point at this stage of the lifecycle is that AI is doing the operational layer of the actual legal work, not just the surrounding business processes. Hours recovered at this stage feed back into capacity, which feeds back into how many clients the firm can take on without adding headcount.

Time saved at this stage: 6 to 8 hours per attorney per week is the conservative number we see at firms running 3 or more of the seven tools. At 2,000 billable hours per attorney per year, that is a 15% capacity gain per attorney without hiring.

Stage 6: Communicate While Working (Status Updates and Client Comms)

Clients leave firms because they feel ignored, not because the legal work was bad. The communication stage is where most attorneys lose otherwise good clients without realizing it.

Status update emails. AI drafts the weekly or biweekly client update based on case activity from the last period. Attorney reviews, edits for tone and any sensitive context, sends. What used to take 30 minutes per client takes 5.

Client portal updates. AI auto generates the portal status post when a milestone hits. Filed motion. Received discovery response. Hearing scheduled. The client sees activity. They feel informed. They do not call asking for status.

Inbound email triage. Tools like Superhuman AI or Microsoft Copilot draft replies to common client emails. Attorney reviews and sends. The 50 emails a day problem becomes the 50 emails a day in 30 minutes problem.

Time saved at this stage: 4 to 6 hours per week per attorney. Plus the harder to quantify but real benefit of higher client retention. Clients who feel informed do not fire their lawyers.

Where this connects to eNZeTi: the same coaching pattern that runs in real time on intake calls runs on follow up touchpoints. Did the staff member confirm the client received the document. Did they set expectations for the next step. Did they ask the right question. We surface the gap and coach the close on every client touch, not just the first one.

Stage 7: Close the Matter (Wrap Up, Reviews, Referrals)

Most firms treat case closure as administrative. The firms growing fastest treat it as a marketing event.

Closing letters. AI generates the standard closing letter populated with case specifics. Attorney reviews and signs. 15 minutes per closure recovered.

Review solicitation. The single highest ROI marketing activity at a law firm is asking happy clients to leave Google reviews. AI sends timed solicitation messages, drafts personalized review request copy, and follows up if there is no response. Firms doing this systematically see 5 to 10x more reviews than firms relying on attorney memory.

Referral asks. AI drafts the personalized referral ask based on the case outcome and the relationship. Specific. Not “if you know anyone who needs a lawyer.” Instead “you mentioned your sister is going through a similar custody situation, would it help if we sent her a free 15 minute consultation as a courtesy.”

Time saved at this stage: 1 to 2 hours per matter closure. Plus 3 to 5x review velocity, which directly improves Google rankings and lead flow at Stage 1.

Where this connects to eNZeTi: the loop closes here. Reviews and referrals at Stage 7 feed back into Stage 1 demand generation. Firms running our model see lead volume from organic and referral channels grow 25% to 50% in the first six months because the closure motion is finally consistent.

Stage 8: Run the Firm (Back Office, Finance, HR, Compliance)

The final stage is everything that keeps the firm running but is not directly client facing. AI here is unglamorous but high leverage.

Billing. AI generates draft time entries from calendar entries, emails, and document activity. Attorney reviews and approves. Captures the 15 to 20% of billable time that gets lost because nobody wrote it down. For a $400 hour attorney billing 1,800 hours a year, that is $108,000 to $144,000 of recovered revenue per attorney per year.

Financial review. AI summarizes monthly P&L, AR aging, trust accounting reconciliation. Partner reads a 1 page summary instead of 20 pages of QuickBooks reports. 4 hours per month recovered at the partner level.

HR and hiring. AI drafts job descriptions, screens applications for fit, generates interview questions tailored to the role. Useful for paralegal and coordinator hires. Less useful for attorney hires because the bar membership and bar standing checks still require human judgment.

Compliance. CLE tracking, conflict alerts on new matter intake, malpractice insurance renewal reminders. AI as a calendar of “things you cannot afford to forget.”

Time saved at this stage: 5 to 10 hours per month at the partner and operations manager level. Plus the recovered billable revenue from better time capture, which is often the biggest financial impact in the entire lifecycle map.

Where this connects to eNZeTi: we surface intake KPIs to the operations manager dashboard alongside billing and matter data. The whole firm runs from one set of numbers instead of three disconnected systems.

The 90 Day Adoption Plan

You just read about eight stages and a dozen specific tools. Most attorneys read this and feel overwhelmed. Do not start with twelve tools across eight stages. Start with one tool in one stage.

Days 1 to 7: Audit the lifecycle.

Map your firm against the eight stages above. For each stage, score yourself 1 to 5 on how AI ready you are today. 1 means you have nothing. 5 means you are running fully automated workflows. Most firms score 1s and 2s across the board. That is normal. The map tells you where the lowest hanging fruit is.

Days 8 to 30: Pick the single highest leverage stage and pilot one tool there.

If your intake conversion is below 30%, start at Stage 3 with eNZeTi. If you are losing 20+ hours a week to discovery responses, start at Stage 5 with Briefpoint. If your PI demands take 80 hours each, start at Stage 5 with Supio. If your billable hour capture is below 75%, start at Stage 8 with AI billing assistance.

Pilot one tool. Run it on a real workflow. Measure the actual time saved or revenue lifted against pre AI baseline. Document what worked and what did not.

Days 31 to 60: Add the second tool.

Now you have one tool in production. The second tool is faster because the operational mindset is already there. The choice is usually one of two patterns. Either you go DEEP in one stage by adding a second tool there, or you go WIDE by addressing a different stage.

Deep example: started with eNZeTi for intake conversion. Add Spellbook for fee agreement generation at Stage 4. Now you have AI from “lead picks up the phone” through “client signs.”

Wide example: started with Briefpoint for discovery at Stage 5. Add Parrot for depositions at Stage 5. Now your whole litigation stage is covered.

Build SOPs around both tools. Anyone who joins the firm uses them by default.

Days 61 to 90: Integrate, train, and roll forward.

This is where firms separate from each other. The firms that win do three things in this window. They make the AI workflow the default and the manual workflow the exception. They train every new hire on the AI tools as part of onboarding. They appoint one person, not the IT person but an operations person, as the AI ops lead with authority to roll out the next tool.

Solo attorneys: pick the one tool that addresses your single biggest time sink. That is your whole 90 day plan. Solo attorneys should not be running pilots on three tools. Pick one. Get the hours back. Reinvest them in client work or in your own time.

Small firms (2 to 10 attorneys): appoint an AI ops lead. Run two tools across two stages by day 90. Aim for 6 to 8 hours per attorney per week recovered.

Mid size firms (10 to 50 attorneys): the AI ops lead is now a part time role for an existing operations manager or a new hire at the operations manager level. Run three tools across three stages by day 90. Aim for capacity gains that defer the next hire.

Large firms (50+ attorneys): AI rollout is a six month project, not a 90 day project. The 90 day version focuses on one practice group as the pilot. Document the playbook. Roll to the rest of the firm in months 4 through 6.

The Strategy Call

The seven tools are public. Anyone reading this article can sign up for any of them today.

The hard part is figuring out which two or three would actually move the needle at your firm, and which stage of the lifecycle to start with. That depends on your case mix, your staffing, your billable rate, your current conversion data, and where the hours are actually being lost. We have run this audit at law firms across multiple practice areas. The pattern is always the same. The two tools that matter for this firm are not the same as the two tools that matter for the firm down the street.

We do this audit on a 30 minute call. No pitch.

Here is what you walk away with:

• Top three AI use cases ranked for your firm by expected ROI
• The specific lifecycle stage where you have the most leverage right now
• Estimated billable hours reclaimed per attorney per week
• A 90 day implementation roadmap on a single page you can show your partners

You can book at https://enzeti.com/calendar/.

We onboard four firms a quarter. There are two slots left for the current quarter at the time of this writing.

If you do not want a call, that is fine. Take the seven tools above and start with the one that maps to your biggest time sink. You will still get most of the value.

If you do want the call, we will tell you which of the seven we would pilot in your firm first, which lifecycle stage to start with, and what is going to slow you down before it slows you down. Most firms do not need seven tools. They need two, deployed correctly, with the SOPs to keep them deployed when the original champion leaves.

That is the playbook.

Charlotin’s database keeps growing. So does the cliff. The attorneys who are pulling ahead are not the ones avoiding AI entirely. They are the ones who learned the cliff, walked along the safe edge, and put AI to work in the eight stages of their firm where it actually pays.

You are reading this. You are already ahead of most.

The next step is one phone call.

https://enzeti.com/calendar/