Call Casey by name.
Get a concise spoken answer first, then ask for the longer explanation.
Back to products← Alchemy Loop portfolio
COMING SOON
CaseyMeeting Expert
Your Personal AI Meeting Expert
For Microsoft Teams & Mobile
Casey joins the conversation with permission, answers from approved knowledge and live operational information, and researches questions while the meeting continues.
Keep the meeting moving
The answer should arrive while the decision can still be made.
A question nobody can answer often becomes a week of research, follow-up and another meeting. Casey brings approved knowledge, current information and outside research into the conversation when the team needs it.
Let the meeting continue.
Casey works in the background and announces when the briefing is ready.
See where it came from.
Company knowledge, live information and public research remain clearly identified.
What Casey does
More than a transcript and summary.
Casey is a subject-aware participant designed to answer, investigate and connect the meeting to the information behind the work.
Answer from approved knowledge
Casey retrieves the relevant parts of your documents and knowledge packs, answers aloud and identifies where the answer came from.
Research while the meeting continues
Ask Casey to investigate a topic. The conversation carries on, and Casey announces when the briefing is ready.
Follow the conversation
With transcription enabled, Casey understands references to earlier discussion and remembers his own answers for the rest of the call.
Join only when invited
Require Casey to wait for his name, or allow him to offer help when a genuine question is left unanswered.
Reach live operational information
Paired with LLModelQuery, Casey can present a governed, read-only answer from connected operational systems.
Keep the record useful
When enabled, meeting notes preserve the discussion, Casey’s answers and the research completed during the call.
One expert, two ways to work
In Microsoft Teams—and in your pocket.
Casey is being prepared for Microsoft Teams meetings and mobile use in rooms, site visits, labs and other work that does not happen on a screen.
Present for the conversation
Follow a consented meeting, answer aloud, carry out background research and prepare notes when transcription is enabled.
- Called by name
- Optional proactive help
- Meeting-aware follow-up
The same expert travels with you
Hold to ask one question or enable continuous listening after reading the consent notice to the room.
- Tap or hold to ask
- Always-listen mode
- One action to stop listening
Tailscale private network. Only your own admitted devices can reach Casey; the service is never open to the public internet.
The mobile experience
Casey travels with the meeting.
Use the main listening screen, start research without stopping the conversation, and open the controls you need for the people and meeting in front of you.
Main meeting screen
Choose the meeting, control when Casey listens and bring Casey into a Microsoft Teams call.
Research during the call
Ask Casey to investigate a topic while the room keeps working, then return when the briefing is ready.
Meeting controls
Open Research, Who’s here and Meeting options only when they are needed.
Consent and control
Participation is a choice—not a background assumption.
Teams decide when Casey can listen, what context is retained, when he may speak and which AI environment handles the work.
Consent first
Casey does not listen until the room has agreed.
Transcript optional
Turn transcription off and Casey receives only the question addressed to him.
Source identified
Answers distinguish approved knowledge, live information and public research.
Participation controlled
Proactive contributions can be disabled so Casey speaks only when called.
Your AI account
Use your own OpenAI or Anthropic business account and provider agreement.
Private model available
Run Casey with a private model when questions and results must remain inside a dedicated AI environment.
Live operational information
Casey never connects directly to your database.
When paired with LLModelQuery, Casey can turn a spoken question into a governed, read-only request against connected operational systems. LLModelQuery runs the query inside your environment; Casey receives the resulting answer and the limited supporting rows needed to explain it aloud and keep the context during that meeting.
The boundary is explicit. LLModelQuery on its own does not send database rows to an LLM. When Casey presents a live-data answer, that specific answer and limited supporting result may be sent to your selected AI account. A private-model deployment keeps that material inside your dedicated AI environment.
MEETING QUESTION→LLMODELQUERY
READ ONLY→LIMITED RESULT→CASEY
SPOKEN ANSWER
READ ONLY→LIMITED RESULT→CASEY
SPOKEN ANSWER
Built from our own work
First built for Alchemy Loop. Now being prepared for customers.
Casey began as an internal product: an expert that could join our calls, travel on a phone, understand our subjects and research what the meeting needed next.
Casey · Coming soon
Bring an expert into the conversation.
Request early access to Casey for Microsoft Teams and mobile.