Convay AI for Physical Meetings

"AI for Physical Meetings" in Convay automates note-taking by generating transcripts and AI-based summaries from audio recordings of in-person meetings. Users start by entering a meeting title and description, record audio, and end the session to create a transcript. They can edit the transcript, rename speakers, use AI to summarize, and export the results.

CATEGORY:

Web Design,

Product Design

ROLE:

UX Designer,

UI Designer,

UX Researcher

TOOLS:

Figma

About Convay

Convay is an innovative video conferencing platform that redefines virtual collaboration by integrating tools for seamless communication and teamwork. Designed with the needs of modern businesses in mind, Convay goes beyond traditional video calls by offering a unified platform that manages all aspects of meeting lifecycles—from preparation and execution to post-meeting follow-ups.


The platform provides enterprise-grade security, ensuring data privacy with end-to-end encryption and secure cloud storage. Key features include high-quality audio and video conferencing, automatic transcription, real-time collaboration tools, and centralized file management through Convay Cloud. Users can organize meeting files, access chat logs, and navigate through media and documents effortlessly. Additionally, Convay supports complex requirements with scalable architecture, making it suitable for organizations of all sizes.


With specialized solutions like AI-powered transcription and summarization, cloud storage for document management, and an intuitive interface for navigating content and activities, Convay aims to create a truly cohesive virtual meeting experience that enhances productivity across teams.


For more details, visit Convay.

Feature Overview

"AI for Physical Meetings" in Convay aims to automate note-taking by generating transcripts and AI-based summaries from in-person meeting audio. The feature was in the research phase, with plans to bring it to market. It allows users to start by entering a meeting title, record audio, and generate a transcript upon ending the session. Users can then edit the transcript, rename speakers, summarize with AI, and export the results.

Problem Statement

In physical meetings, capturing and documenting information accurately can be challenging. Relying on manual note-taking or post-meeting summaries often leads to incomplete or inaccurate records, making it difficult for teams to reference key points or follow up on action items. Additionally, the time-consuming nature of creating meeting notes from scratch poses a risk of missing important details, which can hinder productivity and team alignment. To address these issues, there was a need for an automated solution that could seamlessly record, transcribe, and summarize meeting content, making it easily accessible for all participants.

My Role and Responsibilities

I was involved in the research and early design of the feature. My tasks included designing the user flow, creating the interface, and collaborating with developers. I used feedback from internal testing to improve the design and prepare it for future market release.

Design Process

  • Research: During the research phase, we explored the potential of automating meeting documentation to address the pain points of manual note-taking. We conducted user interviews and analyzed the requirements for accurate transcription and AI summarization.

  • Ideation and Design: I designed the initial user flow to cover key steps, as illustrated below:

Meeting Dashboard Page: Users can access all meetings, start new ones, upload previous recording file or view recent recordings.

Meeting Information Page: Users can enter the title and description and click "Start" or "Cancel."

Main Meeting Panel Page: Featuring a large microphone icon and live transcript display.

End Session Confirmation Modal: Prompting users to confirm ending the session.

Loading UI for Transcription: A visual indication of the transcription process.

Meeting Transcription Page: Where users can view and edit the transcription.

AI Summarization Loading Screen: A loading skeleton while the AI summarizes the meeting.

Meeting Summary Page: Displays the summarized content.

Export Options: For downloading the meeting summary and transcript.

Transcript Editing Options: A 3-dot modal for editing the transcript and renaming speakers.

Edit Transcript Page: Users can make changes to the transcript.

Rename Speaker Page: Allows for renaming speakers.

Edit Meeting Title Page: Users can modify the meeting title.

  • Iteration and Testing: During testing, we developed two versions of the Meeting Panel page to evaluate user preferences:

This version aligns with Convay's main theme, providing a cohesive look.

A lighter interface for improved visibility, but it did not match the overall aesthetic of the Convay platform.

Feedback from users indicated a clear preference for Design A (Dark Mode) due to its consistency with the rest of Convay's design. Additionally, we modified the microphone mute button by removing the text for a cleaner, more intuitive interface. This feedback guided us in finalizing the design for implementation.

Challenges and Solutions

  • Challenge: Ensuring transcription accuracy, especially in noisy environments, was a major issue during the research phase.

    Solution: We integrated advanced noise reduction and speaker identification to improve the quality of the transcriptions.

  • Challenge: Convay's main interface was designed in dark mode, but initial designs for the Meeting Panel were in light mode, creating inconsistency.

    Solution: User feedback indicated a strong preference for the dark mode. We redesigned the Meeting Panel to align with Convay’s overall aesthetic, ensuring a seamless experience across the platform.

  • Challenge: Designing an intuitive editing interface for transcripts that could handle bulk changes, such as renaming speakers.

    Solution: We made the editing tools simple to use, allowing for easy speaker name changes that would update consistently throughout the transcript.

  • Challenge: Making the AI summarization accessible and useful without overwhelming users.

    Solution: We designed a straightforward "Summarize with AI" button, positioning it prominently to encourage use while keeping the feature optional.

Outcome and Impact

Since the feature was still in the research phase, it hadn't yet reached the market. However, internal testing showed promising results. The automatic transcription achieved a high accuracy rate in controlled settings, and testers found the AI summarization useful for quickly reviewing meeting highlights. The simple editing tools also made it easy to correct any errors. These initial successes set a solid foundation for future development, aiming to streamline meeting documentation and improve productivity.

Takeaways

Creating Convay’s design system taught me the importance of balancing structure and flexibility to support both consistency and scalability. Building reusable elements like color variables, typography, and grid layouts allowed for a cohesive design that can grow with the product. I also learned the value of clear documentation and collaboration with developers, which streamlined the workflow and maintained a unified look across features. Above all, this experience reinforced that even the smallest design decisions should enhance the user experience, keeping the system user-centered and future-ready.