Most meeting agendas are blank pages — assembled in a rush, forgotten by the next session. Evro, a communication intelligence platform, is designed to solve exactly this problem by turning your existing meeting history into a ready-built agenda before you walk in the door. This guide covers the DIY AI prompt technique that power users are already using, why it eventually hits a ceiling, and how Evro's Auto Meeting Prep closes the remaining gap automatically.
We have all seen it. Ten minutes before the weekly sync, a Slack notification pings. It is the project lead asking: “Does anyone have anything for the agenda today?”
The silence that follows is not because there is nothing to talk about. It is because everyone is context-switching from their actual work and trying to remember what happened last Tuesday. In a rush, people ping back half-baked items like “Update on Project X” or “Budget discussion.” These are not agenda items. They are placeholders for a conversation that is already halfway finished but has no memory of where it started.
The reality is that most meeting agendas are either non-existent or hastily gathered leftovers. This creates a cycle of repetition that drains the collective energy of the team. But there is a better way — and it does not involve starting from a blank page.
The real agenda for your next meeting is already written. It is buried in the transcripts of your last three sessions.
Part 1: The Problem With Agenda-From-Scratch
The traditional way of building an agenda is fundamentally broken because it assumes each meeting is a standalone event. In reality, work is a continuous flow. When we treat a recurring meeting as a “new” event every week, we fall into the “Status Update Loop.”
The Status Update Loop
You know this loop well. It is the part of the meeting where everyone spends forty minutes saying, “Here is what I did, here is what I am doing, and here is what is in my way.” It feels productive because people are talking, but there is zero continuity. Issues that were “in the way” last week are still there this week, often described in the exact same words.
The reason no one moves past the status update is simple: no one reads the previous transcript. They are too long, filled with “umms” and “ahhs,” and frankly, no one has the thirty minutes required to parse through five thousand words of dialogue before a thirty-minute meeting.
The Data on Disengagement
At Evro, we conducted a survey of 300 knowledge workers to understand how this lack of continuity affects the workplace. The results were telling.
Source: Evro survey of 300 knowledge workers.
When meetings feel irrelevant, it triggers a psychological state known as Active Disengagement. This is not just “boredom.” It is a protective mechanism. When a person feels their most limited resource — their time — is being disrespected, they check out. The stress of that “checked-out” mentality does not disappear when the call ends. It lingers, affecting productivity for the rest of the day.
Psychology Insight
Communication is a learnable skill, but it requires a foundation of shared memory. Without memory, a group cannot develop self-awareness. They just repeat the same patterns until the project ends or someone leaves.
The Elephant in the Room
There is also a social problem with “Agenda-From-Scratch.” The most important topics are often the ones that are the most uncomfortable to bring up. If a project is failing because two departments are not communicating, that rarely makes it onto a Slack-polled agenda. Naming the problem requires someone to have done their homework and be willing to take a social risk. Without a record of past frustrations, it is easier to just talk about “Project X updates” again.
Part 2: The Power User Technique — AI-Assisted Agenda From Transcripts
A small group of “power users” has figured out a workaround that bypasses the blank-page problem entirely. They are using AI not just to summarize a single meeting, but to connect the dots across a series of them.
The Strategy
Instead of asking the team for agenda items, these users feed the transcripts of the last three meetings into an AI assistant like Gemini or Claude. They are not looking for a summary of what was said. They are looking for what was missed.
By looking at a three-meeting window, the AI can see the “slow burn” issues — the topics that appear in meeting one, get deferred in meeting two, and are glossed over in meeting three.
How to Replicate This Today
If you want to try this yourself, here is the architecture of a successful prompt.
Step 1: Collection
Gather the transcripts from your last three meetings with the same group. Most platforms like Zoom, Teams, or Google Meet generate these automatically now.
Step 2: Verification
Quickly check the speaker labels. If the AI thinks “John” is “Sarah,” the logic of who is responsible for what will fall apart. You do not need a perfect transcript, but you need accurate speaker identification.
Step 3: The Prompt
Paste the text of all three transcripts into your AI tool and use a prompt designed to find the gaps.
The Power User Prompt
I am providing transcripts from our last three meetings with the same group. Analyze them as a continuous thread, not as separate meetings. Identify the 3 most important agenda items for our next meeting based only on issues that remain unresolved across these sessions.
Look for:
- deferred or half-made decisions
- incomplete action items or unfulfilled commitments
- recurring blockers or tensions
- topics that repeatedly surface without resolution
- issues that seem important but were softened, avoided, or pushed aside
For each recommended agenda item, return:
- Agenda item title
- What happened previously, in 2 to 3 sentences
- Evidence that it is still unresolved
- Why it matters now
- Likely owner or people who need to weigh in
- The decision or outcome the team should reach in the next meeting
- A suggested question the meeting leader should ask to move it forward
After that, produce:
- A prioritized top-3 list
- A concise next-meeting agenda
- A short “open commitments to revisit” section listing any promised follow-ups that still appear incomplete
Why This Changes the Room
This technique works for three main reasons:
- It Forces Evolution: The meeting is forced to build on the past. You cannot ignore a three-week-old action item when the AI has put it at the top of the list with a “Still Unresolved” tag.
- It Neutralizes Conflict: If a project is stalled, the AI can name that tension neutrally. It is no longer “Bob calling out Alice.” It is “The data shows this topic has been deferred twice.”
- It Surfaces Patterns: It identifies things that humans miss. Maybe every time a certain budget item comes up, the conversation shifts to a different topic. The AI will catch that pattern and ask why.
If a project is stalled, the AI can name that tension neutrally. It is no longer one person calling out another — it is the data speaking.
Part 3: The Limitations of the DIY Version
While the manual AI method is a massive leap forward, it eventually hits a ceiling. If you are a manager with six or seven recurring meeting threads, the DIY approach becomes another chore on your to-do list.
The Friction of Manual Prep
To do this right, you have to remember to download the transcripts, clean them up, and run the prompts every single time. It takes about fifteen minutes per meeting. If you have five such meetings a week, you are spending over an hour just on “prep for the prep.” Most people eventually stop doing it because the friction is too high.
The Context Gap
Standard AI assistants also have a “flat” understanding of your world. They treat every speaker the same. They do not know that “Sarah” is the stakeholder who actually makes the final call, or that “David” tends to be overly optimistic about deadlines. Without that relational context, the suggested agenda can sometimes feel technically correct but socially tone-deaf.
How Evro Approaches This
This is exactly the problem Evro’s Auto Meeting Prep feature was built to solve — removing the manual labour entirely and adding the relational intelligence that generic AI tools do not have access to.
Pre-Meeting · Auto Meeting Prep
The Brief Builds Itself
Every time a meeting appears on your calendar, Evro starts working in the background without you doing anything. It reviews your full interaction history with each attendee — not just the last three meetings, but every meeting you have had with those people since you joined Evro. By the time the meeting starts, your brief is ready: a concise summary of where the last session ended, the decisions that were made, the ones that were deferred, and the threads that were never properly closed.
Pre-Meeting · Commitment Tracking
Nothing Slips Through
Every action item assigned to anyone in the room — whether they said “I’ll get back to you by Friday” or “Let’s circle back on that next week” — is surfaced before you walk in. If something was promised and has not been marked done, it appears in your brief as an open commitment. The meeting does not start from zero.
Pre-Meeting · Communication Guidance
Context-Specific Guidance for the People in the Room
Evro has been reviewing your meeting history and building a picture of how each person communicates — what kind of arguments land with them, whether they prefer direct asks or context first, where they tend to disengage or push back. Auto Meeting Prep uses that to give you specific, practical suggestions for how to land your key messages with those people in that meeting — not generic tips about “active listening,” but context-specific guidance grounded in your actual interaction history with them.
In-Meeting · Real-Time Guidance
A Private Second Set of Eyes
During the meeting itself, Real-Time Meeting Guidance can provide optional, private nudges visible only to you — a quiet reminder to pause and check for alignment, a note that you have spent most of the time talking and it might be worth opening the floor, or a prompt that you are running low on time and the key decision still has not been surfaced. No one else sees them.
The Relational Layer Is the Differentiator
A standard AI assistant treats everyone in a transcript the same. It does not know that your stakeholder Sarah is the person who actually owns the budget decision, or that David tends to raise concerns about timelines late in the conversation once he has had time to think. Evro is currently building an About Others feature that builds an evolving profile of each person you work with regularly — their communication style, their patterns, where your style creates friction with theirs, and how the relationship is trending over time.
On Privacy
Communication feedback and relational profiles in Evro are visible only to you — they are never shared with managers, teams, or other attendees. The intelligence is private by design.
How It Compares
The manual method gets you 30% of the way there. Evro closes the rest — and does it without adding a task to your already overloaded prep routine.
What this guide establishes
01
Most meeting agendas fail because they treat each meeting as a standalone event — severing the continuity that makes recurring meetings productive
02
The DIY AI prompt technique is a significant upgrade, but breaks down at scale and lacks the relational intelligence needed to be truly useful
03
Evro’s Auto Meeting Prep removes the manual work entirely and adds the relational layer — surfacing what matters, for the specific people in the room, every time
Frequently Asked Questions
Can I use any AI tool to create a meeting agenda from transcripts?
Yes — any large language model like ChatGPT, Claude, or Gemini can process transcripts and identify unresolved items if you paste them in manually. The quality of the output depends heavily on the quality of the prompt. The limitation is that this approach requires manual effort every time and the AI has no understanding of who the speakers are or how they typically behave.
How many transcripts should I include in the AI prompt for best results?
Three is the practical sweet spot. One transcript gives you a summary; two lets you compare; three is where patterns start to emerge. You can include more, but most AI tools have context window limits, and the added value diminishes past three to four sessions unless there is a specific long-running issue you are tracking.
What is the difference between a meeting summary and a meeting agenda?
A meeting summary is a record of what happened. A meeting agenda is a structured plan for what should happen next. The DIY technique and Evro both focus on agenda generation — taking the unresolved items, deferred decisions, and open commitments from past sessions and turning them into the forward-looking structure for the next one.
What does Evro’s Auto Meeting Prep include?
Auto Meeting Prep builds a pre-meeting brief automatically before each calendar meeting. It includes a context continuity summary (where the last session ended), a commitment tracker (open action items from any attendee), and communication guidance specific to the people in the room — based on Evro’s ongoing analysis of your interaction history with each person.
Is the communication guidance in Evro visible to my manager or teammates?
No. Communication guidance, relational profiles, and pre-meeting briefs in Evro are visible only to you. They are never shared with managers, other attendees, or your organization. The intelligence is private by design.
Get Started with Evro
Stop building your agenda from scratch
Evro’s Auto Meeting Prep works in the background so your brief is ready before the meeting starts — with the relational context that generic AI tools can’t provide.
Explore EvroSubscribe to our newsletter
Stay up to date on our product updates,
latest news, upcoming events and more.