The EOS Level 10 meeting is one of the most structured business meetings in existence. Ninety minutes, same agenda every week, same roles, same expectations. It’s designed to produce consistent results through consistent execution.

Yet most EOS-running companies still rely on a human note-taker who’s supposed to participate AND document simultaneously. That person either takes great notes and barely contributes, or contributes fully and produces weak notes. Neither outcome serves the team.

The structure that makes Level 10 meetings effective also makes them perfect for AI capture. Every section has a distinct purpose. Every conversation follows a predictable pattern. The most valuable content — the Issues Solving section — is exactly the content that gets lost when note-taking is an afterthought.

Our AI meeting recap guide covers how AI captures meetings in general. This post is specific to EOS Level 10 meetings: what AI extracts from each section, how to build week-over-week intelligence, and why L10 meetings are the ideal use case for AI-powered documentation. For broader context on documentation practices, see our guide to meeting documentation best practices.

The L10 Structure: What AI Extracts from Each Section

A standard Level 10 meeting follows the same agenda every week. Here’s what happens in each section and what AI captures.

Segue: Personal and Business Bests

The meeting opens with each participant sharing a personal best and a business best from the past week. This two-minute exercise builds team connection and sets a positive tone.

What AI captures: Each person’s shares, attributed to them individually. Over time, this builds a relational map of your team — who’s celebrating personal milestones, what business wins are being recognized, how team sentiment trends from week to week.

The Segue data is often overlooked in the moment but becomes valuable over quarters. When you look back and see that a team member’s personal bests shifted from “family vacation” to “nothing really” over six weeks, that’s a signal. AI captures it without anyone needing to track it consciously.

Each weekly metric is reviewed — on track or off track. The team discusses any off-track metrics and identifies root causes.

What AI captures: Every metric, its status (on/off track), the discussion around off-track items, and any root causes identified. AI structures this as a metric-by-metric report with annotations.

Over multiple weeks, this becomes a trend log. “Revenue has been off-track for four consecutive weeks.” “Customer satisfaction moved from green to yellow two weeks ago and hasn’t recovered.” These trends are obvious when you have structured data. They’re invisible when metrics are discussed verbally and recorded as a checkmark in a spreadsheet.

Rock Review: On Track or Off Track

Each quarterly Rock (the 3–7 most important priorities) gets a status check. Is it on track for completion this quarter, or not?

What AI captures: The status of each Rock, any discussion about at-risk Rocks, and specific commitments made to get off-track Rocks back on track. AI links Rock status across weeks so you can see the trajectory — a Rock that was on track in Week 2, flagged in Week 5, and off track by Week 8 tells a story that a single status checkbox doesn’t.

Customer and Employee Headlines

The team shares headlines about customers and employees — good news, concerns, and items to celebrate or address.

What AI captures: Each headline, categorized as customer or employee, tagged as positive or concerning. Over weeks, patterns emerge: repeated concerns about the same customer, a cluster of employee headlines around a specific team, positive trends that deserve recognition.

This section is where relationship intelligence starts. The customer headlines are early warning signals — positive or negative — about accounts that may need attention. AI captures them alongside the rest of the meeting data, creating a connected record.

To-Do List Review: Completed Items

The team reviews the running To-Do list, checking off completed items and noting what’s still open.

What AI captures: Which To-Dos were completed, which remain open, and how long each item has been on the list. AI can flag To-Dos that have appeared in multiple consecutive L10s without completion — a signal that the item either needs to be prioritized or removed.

The To-Do section is deceptively important. In EOS, To-Dos that don’t get completed become Issues. Issues that don’t get solved become patterns. AI captures this chain at every stage.

IDS: Identify, Discuss, Solve — The Most Valuable Section

This is the core of the Level 10 meeting and the section where the most valuable content is generated. The team identifies issues, discusses root causes, and solves them. IDS typically gets 30–45 minutes of the 90-minute meeting.

What AI captures: Every issue raised, the discussion around each one, the solution agreed upon, and the action items that result. This is the richest content in the meeting — and it’s exactly the content that gets lost when note-taking is manual.

IDS conversations are fast, sometimes contentious, and always valuable. The note-taker is pulled between participating in the problem-solving and documenting it. AI eliminates that trade-off. The team focuses on solving. AI focuses on capturing.

Specifically, AI extracts from IDS:

  • Issues identified — what problems were raised, by whom
  • Root cause discussion — what the team determined was the actual cause vs. the symptom
  • Solutions agreed upon — what was decided, including specifics that would be lost in a brief summary
  • Action items — who committed to what, by when
  • Unresolved issues — items that were raised but tabled for a future meeting

Conclude: Recap and Cascading Messages

The meeting wraps with a recap of what was decided and messages that need to cascade to other teams or levels in the organization.

What AI captures: The cascading messages — what information needs to be shared, with whom, and by when. These messages are critical for organizational alignment and frequently get lost between the L10 meeting and the communication actually happening.

The L10 Intelligence Layer: Week-Over-Week Tracking

Individual L10 meeting recaps are useful. The real value emerges when you link recaps across weeks into what I call the L10 Intelligence Layer — a persistent, structured record of everything your leadership team has discussed, decided, and committed to.

How the Intelligence Layer Works

Each week’s L10 generates a structured recap. When these recaps are linked to the same team and time-ordered, you get:

Issue tracking across weeks. An issue raised in Week 3, partially discussed in Week 4, and solved in Week 6 has a complete record. You can see how the understanding evolved, what solutions were tried, and what finally worked. This is far more valuable than a simple “resolved” status.

Rock trajectory. Each Rock’s weekly status builds a quarter-long narrative. You see exactly when a Rock started slipping, what was discussed, and whether the interventions worked. This is retrospective gold when you’re evaluating quarterly performance.

Scorecard trends. Every metric’s weekly status, aggregated over 13 weeks, produces a trend line that no manual tracking system sustains. “Revenue was on track for Weeks 1–5, off track for Weeks 6–10, then recovered after the pricing adjustment in Week 8.” That story exists in your meeting data. AI surfaces it.

To-Do completion rates. How quickly does the team complete its To-Dos? What percentage carry over from week to week? Are certain types of To-Dos consistently late? These metrics reveal execution patterns that affect every aspect of the business.

Team participation patterns. Who raises issues? Who volunteers for To-Dos? Who follows through? Over a quarter, these patterns paint a picture of team dynamics that’s useful for leadership development and accountability.

Why This Matters for EOS Implementers

The L10 Intelligence Layer turns every Level 10 meeting from a discrete event into a data point in a continuous record. For EOS Implementers — the consultants and coaches who help companies adopt EOS — this is a differentiator.

When you can show a client their 13-week issue resolution rate, their Rock completion trajectory, and their scorecard trends — all captured automatically from their existing L10 meetings — you’re not just facilitating. You’re providing data-driven coaching. The client sees the impact of EOS in their own data, not in anecdotes.

Why L10 Meetings Are Perfect for AI Capture

Not all meetings benefit equally from AI. L10 meetings are an ideal use case for three specific reasons.

They’re Highly Structured

AI works best when it can categorize information by section. The L10 agenda provides those sections naturally. The AI knows that discussion in the first 5 minutes is Segue, discussion about metrics is Scorecard, and the longest section is IDS. This structure makes the AI output cleaner and more useful than it would be for an unstructured meeting.

They Recur Weekly

AI value compounds over time. A single meeting recap is useful. Twelve weeks of meeting recaps, linked and structured, are powerful. The weekly cadence of L10 meetings means the AI system builds a rich dataset within a single quarter. Compare this to monthly strategy meetings, where it takes a year to accumulate similar data volume.

The Most Important Content Is the Most Likely to Get Lost

IDS is where problems get solved. It’s also the section where note-takers are most engaged in the conversation and least able to document. The emotional intensity of problem-solving, the rapid back-and-forth of discussion, and the complexity of root cause analysis all work against manual note-taking. AI captures it without effort from any participant.

Setting Up AI Capture for Your L10 Meetings

The implementation is straightforward. Connect your AI meeting tool to the calendar event for your weekly L10. The tool joins the meeting, records the conversation, and produces a structured recap within minutes.

For best results:

  • Use a consistent meeting name. “L10 — Leadership Team” or “Weekly Level 10” helps the AI tool organize recaps into a single series.
  • Let the Integrator know the tool is active. The person running the meeting should know AI is capturing. This isn’t about surveillance — it’s about ensuring the team is aware and comfortable.
  • Review the recap after each meeting. Spend 2–3 minutes confirming that issues, To-Dos, and decisions were captured accurately. This builds trust in the system and catches errors early.
  • Link recaps to your EOS documentation. If you use an EOS tool or shared document for Rocks, issues, and To-Dos, reference the AI recaps as the source of truth for what was actually discussed.

FAQ

Does AI capture work for same-room L10 meetings, or only video calls?

Most AI meeting tools are designed for video calls (Zoom, Google Meet, Teams). For same-room meetings, you can use a laptop with the AI tool active and a decent microphone. Audio quality is lower but still produces useful recaps. Some teams run a hybrid setup where one participant joins from their laptop even though everyone is in the same room.

What if our L10 meetings run longer than 90 minutes?

AI tools handle meetings of any length without issue. However, if your L10s consistently run over 90 minutes, that’s a process issue worth addressing. The most common causes: too many issues in IDS, spending Scorecard time on problem-solving instead of flagging, or insufficient preparation before the meeting. The AI data can actually help you diagnose this — if you see that IDS consistently takes 60+ minutes, it’s a signal to tighten issue selection.

How do we handle confidential issues during IDS?

Most AI tools let you pause recording during specific sections. If an issue involves personnel decisions, compensation, or other sensitive topics, the Integrator can pause the AI capture for that discussion and resume afterward. Alternatively, you can set rules to exclude certain types of meetings or delete specific sections of the recording after the fact.

Can we use AI capture alongside our existing EOS tools?

Yes. AI capture is complementary to tools like Ninety, Traction Tools, or EOS One. Those tools manage the structured data — Rocks, issues, To-Dos, scorecard metrics. AI captures the conversation around that data — the context, nuance, and decision-making process that structured tools don’t record. The two together give you the complete picture: what was decided and why.

RecapCRM captures every section of your EOS Level 10 meeting automatically — Scorecard metrics, Rock status, IDS solutions, and cascading messages — building week-over-week intelligence that compounds across quarters. Start free and see what your L10 meetings have been missing.