I Built a Meeting Agent That Remembers What People Promise

Krishna Veni Ch — Tue, 29 Sep 2026 15:48:45 +0000

Meetings are full of small details that become important later.

Someone says, “I’ll send the revised pricing tomorrow.” 📄
Someone mentions a concern.
Someone shares a preference.
A decision is made.

But after the meeting ends, these details can easily get buried in notes, chats, emails, and transcripts.

That’s the problem we wanted to solve with MeetMind — Your Relationship Memory Agent. 🤝

🚨 The Problem

Before an important meeting, we often spend time trying to remember what happened in the previous conversation.

The information may exist somewhere, but the context is scattered.

📂 Important context is spread across different sources.
🤝 Promises and follow-ups can easily be forgotten.
⏱️ Preparing for the next meeting takes unnecessary time.
💡 Our Approach

Instead of treating every meeting as an isolated conversation, MeetMind gives the relationship a persistent memory.

The workflow is:

Meeting Information → AI Extraction → Persistent Memory → Meeting Brief → Feedback → Improved Memory

Users can provide meeting notes, transcripts, or email summaries. MeetMind extracts important information such as promises, concerns, decisions, and preferences and uses that information to prepare for future conversations.

📌 Promise & Follow-up Tracker

One of our core features is the Promise & Follow-up Tracker.

For example:

“I’ll share the revised pricing tomorrow.”

MeetMind can recognize this as a commitment and track the owner, due date, and status.

Before the next meeting, unresolved or overdue commitments can be brought back into focus.

So an important promise doesn't simply disappear inside a long meeting transcript. 🔎

👤 Stakeholder Intelligence

MeetMind doesn't just remember the meeting — it remembers the relationship.

The Stakeholder Intelligence Card brings together:

🕐 Last interaction
💬 Preferences
📌 Open commitments

This gives the user a quick understanding of what happened previously and what still needs attention.

📝 Smart Meeting Brief

When an upcoming meeting is created, MeetMind uses the stored context to generate a Smart Meeting Brief.

It can include:

🎯 Meeting objectives
📌 Unresolved commitments
⚠️ Stakeholder concerns
📋 Recommended agenda
💬 Suggested opening

Instead of starting preparation from a blank page, the user starts with context from previous conversations.

🧠 Where Hindsight Comes In

This is the most important part of MeetMind.

We use Hindsight as the persistent memory layer behind the agent.

Useful information from conversations is retained as relationship context — including meetings, promises and stakeholders — and can be recalled when generating future meeting briefs.

So MeetMind isn't simply:

Input → Summary → Done

It's designed more like:

🔄 Remember → Prepare → Meet → Learn → Remember Again

The architecture includes the frontend, API layer, AI extraction, persistent memory and brief generator, with persistent memory connecting previous information to future preparation.

🔄 Learning From Meeting Outcomes

After a meeting, the user can answer three quick questions about the outcome, including what happened and what worked or didn't.

That outcome goes back into memory.

Over time, this allows future meeting preparation to become more personalized. 🚀

💭 What We Learned

One of the biggest things we learned while building MeetMind is that an AI meeting assistant doesn't only need to understand what was said.

It needs to remember what matters.

A full transcript can contain thousands of words, but a few details may be much more important:

What was promised?
What matters to this person?
What is still unresolved?
What should happen next?

That's where persistent memory becomes valuable.

🚀 Conclusion

MeetMind started with a simple idea:

What if your meeting assistant could actually remember?

By combining AI extraction with persistent memory, MeetMind connects past conversations with future preparation.

Instead of walking into every meeting and trying to reconstruct the past, users can start with the context they already built.

🧠 Remember. Prepare. Meet. Learn.

That's MeetMind. 🤝

DEV Community: Krishna Veni Ch

I Built a Meeting Agent That Remembers What People Promise