Feature Requests

First, I would like to be able to export a topic or multiple topics. To break it down: When I select a topic to export, I want to be able to perform a bulk export of multiple sessions within that topic. When I export a topic, I should be able to select either all or some of the sessions within that topic. Even within the sessions, I want to be able to drill down and select whether to export the entire data, just the summary, just the to-dos, or the detailed notes. Therefore, all of these options should be available for export per session and per topic.

This proposal suggests extending Hedy with an objective-aware layer that actively tracks meeting outcomes. Objectives definition: objectives are set by the user, selected from predefined meeting types, or inferred from previous meetings and open follow-ups. Live tracking: during the meeting, Hedy links spoken content to objectives, detects decisions and objections, and measures objective coverage in real time. Visual dashboard: a live session view displays each objective’s status (achieved / partial / not covered) with key signals and timestamps. Continuity: after the meeting, objectives are updated with outcomes and carried forward if unresolved. Goal: move Hedy from passive note-taking to an execution-focused meeting copilot.

In addition to the current basic local Whisper model, it could be added Parakeet as ASR model option, which offers excellent performance with low-latency, especially for European languages, the primary user base of Hedy AI. Parakeet is highly efficient, requires only 1GB of storage and runs well on mid‑range devices manufactured in/since 2024, making it a valuable upgrade for improved local transcription quality, even offline. Note: An optimized speaker-identification workflow can be envision with Parakeet generated transcripts being sequentiallly processed by Sortformer, Nvidia's state of the art model for speaker diarization . Currently, Sortformer v2.1 identifies up to 4 speakers, neverthless an 8-speaker version , better suited for meetings, is in progress with an ETA in Q2 2026

Option to automatically end session based on key words recognised like "thank you all" or " that's all" or "end the session". It can even be session automatically ends when there is silence for some duration of time or when there is no relevant inputs coming

Integration with Apple Reminders would be amazing and/or a centralized location to see all to-dos (sort-/filterable by open/closed/date, etc) would be wonderful!

I understand that in order to transcribe the meeting in real-time the transcript is lower quality than as if it was processed as a whole audio session all at once. Perhaps after the meeting I can have the option to re-process the audio from the meeting and get a higher quality transcript?

Would it be possible to expand Hedy to become a speech-to-text asisstant, and let us dictate text anywhere in our apps ? Similarly to https://wisprflow.ai/

This could be processed and added after the end of the session. Ex: Person 1, Person 2, Person 3 etc...

Having a knowledge base where we could upload documents that Hedy uses for custom prompts during meetings would help a lot. I am a data privacy professional so having legislation such as GDPR and the multitude of guidelines from Authorities within the knowledge base would allow specific recommendations following specific frameworks. The same with coaching for example. Having my own framework and process uploaded in a knowledge base would allow Hedy to be a specialised assistant in a meeting and suggest actions based on that particular framework. I am aware that this is somewhat solved by the context window and that there is an initiative for improving it, but a knowledge base goes deeper into customisation. Maybe a hybrid of the 2 could be set up for different flows.

Possibility to search, verify, respond to Hedy's content using other artificial intelligence models (integration like Zemith can do, but within Hedy AI recordings).