Every day, legal requests arrive in your inbox, Slack channels, hallway conversations, and forwarded emails. By the time a request reaches the right lawyer, it has bounced through three people and two days have passed. Legal intake automation solves this. It consolidates requests from every channel into a single form, categorizes them automatically, routes them to the right owner, and tracks them to completion. The result: response time drops from days to hours, and your team can actually focus on legal work instead of request management.

Key takeaways

  • Ad-hoc intake (email, Slack, spreadsheets) breaks down at scale. A centralized intake form is your first step.
  • A five-stage intake framework (receive, classify, prioritize, route, track) lets you design processes that scale.
  • AI-assisted intake categorization and chatbot FAQ responses are quick wins that reduce human triage time by 30-50%.
  • SLA tracking (response times for different request types) creates accountability and visibility.
  • Automation does not mean eliminating human judgment; it means eliminating busywork.

Why ad-hoc intake fails

Most in-house legal teams operate on an intake model that looks like this: requests arrive via email, Slack, Teams, phone calls, and sometimes a shared spreadsheet. Someone (often the GC or a paralegal) acts as a queue manager, reading requests, mentally categorizing them, and forwarding them to the right person. Responses are tracked in someone’s mind or an Excel file. Follow-ups happen when a stakeholder asks “where is my legal review?”

This model works until it does not. Once you reach about 15-20 requests per week, it collapses. Requests get lost. Priorities are unclear. Response times become unpredictable. Your team spends more time managing requests than answering them.

The cost is real. A legal team spending 5-10 hours per week on intake management is losing 25% of their time to triage and communication before any legal work happens.

The five-stage intake framework

A modern intake process has five discrete stages. Automating each stage unlocks efficiency.

Stage 1: Receive

Requests come from everywhere. Instead of trying to police where requests originate, build a multi-channel intake system that funnels everything to one place.

Channels to support:

  • Web form: A simple form on your company intranet or via your CLM (Ironclad, Juro, etc.). This is your default.
  • Email alias: info@lexfolks.com, which routes to your intake form or system of record.
  • Slack integration: A Slack bot that captures requests from your legal Slack channel (e.g., /request-legal-review).
  • Teams or email forwarding: If your company uses email heavily, support info@lexfolks.com forwarding to your form.
  • Chatbot FAQ: For common questions (What is our IP policy? How do we handle data requests?), a chatbot can answer immediately without manual triage.

The goal is simplicity: any way the requester thinks to contact you should land in your intake system.

Stage 2: Classify

Once a request is received, it needs to be categorized. This is where automation shines.

Manual classification: Your paralegal or ops person reads the request and tags it (NDA, vendor agreement, IP, employment, compliance, etc.).

AI-assisted classification: If you use a tool with AI (Juro, Streamline AI, or a custom integration with GPT-4), the system can read the request and suggest a category. Your team reviews and confirms, iterating the model over time.

Classification matters because it drives the next step: prioritization and routing.

Stage 3: Prioritize

Not all legal work is equally urgent. Requests need to be tiered.

A simple triage system:

  • Red (Urgent): Litigation, regulatory response, deal-related work. Response time: today or next morning.
  • Yellow (Standard): Vendor agreements, partnership contracts, employment issues. Response time: 2-5 business days depending on complexity.
  • Green (Routine): NDAs with counterparties you know, internal policy questions, general counsel. Response time: 5-10 business days.

Prioritization can be automated: classify the request, then assign a tier based on work type and urgency flags the requester provides.

Stage 4: Route

Once classified and prioritized, the request needs to go to the right person.

Routing rules:

  • NDAs go to the paralegal who manages vendor relationships.
  • Employment questions go to your employment counsel (in-house or fractional).
  • Vendor agreements go to your GC.
  • Compliance questions go to your Chief Compliance Officer or external counsel.
  • Routine IP questions go to your ops team or are answered by a chatbot.

Routing can be manual (your ops person assigns it) or automated (your tool applies rules based on category and complexity).

The insight here: even if a lawyer has to do the work, a non-lawyer can ensure it lands on the right desk the first time. This saves the back-and-forth that wastes days.

Stage 5: Track

Every request needs a status: new, in progress, pending input, done, escalated.

A simple matter tracking view shows:

  • Request ID
  • Requester and their department
  • Work type and priority
  • Assigned owner
  • Current status
  • Deadline
  • Days since receipt (to flag SLAs)

This view lives in your CLM, matter management tool, or even a shared spreadsheet with conditional formatting. The point is visibility: your team can see what is incoming, your GC can prioritize, and you can spot bottlenecks.

Automation opportunities

Once you have the framework in place, here are the highest-value automation plays.

Intake form routing

If your form is in Juro, Ironclad, or a similar CLM, or even in Zapier, you can use conditional logic to route requests automatically. Example: if work type is “NDA,” assign to the paralegal and set SLA to 2 days. If work type is “employment,” assign to your fractional employment counsel and set SLA to 5 days.

Effort: Low. Most tools support basic conditional routing. Payoff: eliminates manual assignment, reduces delay by 50%.

AI triage and categorization

Feed the request body into a language model (GPT-4, Claude, or a legal AI like Westlaw AI-Assisted Research). The model reads the request, suggests a category, and flags urgency.

Example: “Our VP Sales just told me we need to sign a vendor agreement with Acme Inc. by EOD tomorrow” gets classified as vendor agreement + red priority automatically.

Effort: Medium. Requires API integration or a tool that supports it natively. Payoff: eliminates manual reading and categorization, 40-50% time savings on triage.

Chatbot FAQ responses

For the 20-30% of requests that are frequently asked questions, deploy a chatbot. Example questions:

  • “What is Lexfolks’ IP policy?”
  • “How do we handle data subject access requests?”
  • “What is our data retention policy?”
  • “How do we approve new SaaS tools?”

A chatbot (ChatGPT, your company’s internal bot, or tools like Intercom) can answer these immediately without a lawyer’s involvement. If the requester still needs legal review, the request flows to your intake system.

Effort: Low to medium. Most platforms support chatbots. Payoff: 30-40% of requests are resolved without touching your team.

SLA dashboard and escalation

Create a dashboard that shows response time for each request. If a request exceeds its SLA (e.g., a “standard” vendor agreement is overdue after 5 business days), it automatically escalates to your GC.

Effort: Medium. Requires a tool that tracks intake metrics. Payoff: prevents bottlenecks, ensures accountability, eliminates stakeholders asking “where is my request?”

Feedback loop and learning

After each request is resolved, send a simple survey to the requester: “Did we meet your timeline? Was the answer helpful? Any feedback?” Use responses to refine your SLAs and processes.

Effort: Low. Most forms tools support post-completion surveys. Payoff: continuous improvement, data for explaining changes to your team.

Building your first intake system: A step-by-step start

Step 1: Audit current state (Week 1)

Track every request that comes in for one week. Where does it come from? How long does it take to answer? Where does it get lost? This baseline justifies investment and identifies quick wins.

Step 2: Design your intake form (Week 2)

Create a form (Google Form, Typeform, Jotform, or your CLM’s native form). Ask:

  • Requester name and department
  • Brief description of the request
  • Work type (drop-down: NDA, vendor agreement, employment, compliance, IP, other)
  • Deadline (when do you need this?)
  • Supporting documents (optional)
  • Urgency (low / medium / high)

Keep it to 5-7 fields. Longer forms reduce completion rates.

Step 3: Choose your system of record (Week 2-3)

Where will requests live? Options:

  • Simple: Google Sheets with a form feed. Free, low friction, but limited automation.
  • Mid-market: Juro, Ironclad, or your existing CLM. Native integration with contract workflows, e-signature, and playbooks.
  • Custom: A Zapier or Make workflow that feeds requests into your tool of choice (Airtable, Asana, Monday).

For most teams, a CLM or matter management tool is worth the investment.

Step 4: Set up routing and SLAs (Week 3)

Define: which request types go to whom, and what is the expected response time for each tier?

Example SLAs:

  • Standard NDA: 2 business days
  • Standard vendor agreement: 5 business days
  • Complex vendor agreement: 10 business days
  • Employment question: 2 business days
  • Routine IP question: 5 business days
  • Litigation or regulatory: next business day

Build these into your system. Set escalations for overdue requests.

Step 5: Communicate and train (Week 4)

Send an email to your company explaining the new intake process. Include: the form link, what to do if they have questions, and expected response times by request type. Do a brown bag or Slack announcement. Address the most common question: “What if I email the GC directly?” Answer: “Your request will be rerouted to the intake form so it does not get lost. This ensures you get an answer on time.”

Step 6: Monitor and refine (Weeks 5+)

Check your intake dashboard daily for the first month. Are requests being routed correctly? Are SLAs being met? What types of requests are bottlenecking? Weekly, touch base with your team: what is working, what is not?

AI guardrails for intake

If you use AI for triage or chatbot responses, remember: AI is a speedup, not a replacement. Always gate AI-assisted categorization with human review, at least initially.

For chatbots answering FAQs, ensure:

  • Accuracy: The chatbot is only answering pre-approved questions with known answers.
  • Escalation: If the bot is uncertain, or if the requester indicates they need more than an FAQ answer, escalate to a lawyer.
  • Audit trail: Log what the bot answered and when. This provides protection if there is a dispute.

See /resources/ai-legal-ops-hallucination-risk for more on AI guardrails in legal workflows.

Measuring intake success

Track these metrics weekly or monthly:

  • Intake volume: How many requests are coming in?
  • Intake-to-assignment time: How long between receipt and routing to the right person? Target: same day.
  • Assignment-to-response time: How long from assignment to first substantive response? This varies by work type, but should align with your SLAs.
  • SLA compliance rate: What percentage of requests are getting initial responses within their SLA? Target: 90%+.
  • Cycle time by work type: For each category (NDA, vendor agreement, etc.), how long end-to-end? Track trends over time.
  • Requester satisfaction: Simple survey or NPS. Are people satisfied with response time and quality?

Scaling beyond intake

As intake becomes automated, you unlock capacity for the next layer of operations: standardizing contract review through /resources/contract-playbook-standardized-terms, implementing spend management (see /resources/legal-spend-management-outside-counsel), and building a KPI framework (see /resources/legal-kpis-metrics-framework).

FAQ

Can we automate intake if we do not have a CLM?

Absolutely. A Google Form feeding into a Google Sheet with conditional formatting and email notifications will cover basic intake. As you grow, you can migrate to a CLM, but you do not need one to start.

How do we handle requests from the CEO or board?

Set an exception: mark these as red priority and route immediately, bypassing the queue. But still log them in your intake system so you have a full record.

What if our team rejects the intake form and keeps emailing the GC directly?

Resistance is normal. Address it with a soft deadline. For the first month, your ops person can gently redirect emails to the form. After a month, the GC stops responding to email requests (except urgent ones) and directs people to the form. This sounds harsh, but it works.

Should we use AI for all triage, or keep some manual?

Hybrid is best. Use AI for initial categorization and triage, but have your paralegal or ops person review and confirm, especially for the first 100 requests. As the model learns your types and patterns, gradually reduce manual review.

How long does it take to see ROI on intake automation?

You should see immediate gains: intake time drops in week one, response time improves in week two. Over a quarter, you will see 30-50% reduction in time spent on intake management and 40-60% improvement in average response time.

Legal intake automation is foundational. You cannot optimize the rest of your legal operations until requests are flowing through a predictable system. If you are managing requests via email and spreadsheets, consolidating intake should be your next move. We have helped dozens of in-house teams and law firms build intake systems that scale. Contact us to discuss how we can help you design and implement intake automation.