Reference & expert calls
Reference and expert calls handles the part of diligence that is mostly organising. It proposes who is worth speaking to and why, takes in the transcripts and notes once the calls have happened, and aggregates the signal across every conversation on the deal. What the deal team wants is the pattern across calls, and this puts it in one cited place.
Step by step
Write down the diligence question
The team states what the calls are meant to settle. Every later step is judged against that question.
Identify and rank candidates
The agent proposes people to approach from your own network files and the sources you have approved, ranked by how close they sit to the question.
Ingest transcripts and notes
Transcripts, interview notes and any written follow-up are read in from the folder your team drops them into.
Aggregate the signal across calls
Points that recur across conversations are grouped, each attributed to the call it came from.
Surface where sources conflict
Where two people said different things, the agent presents both with attribution rather than averaging them into one line.
Associate approves before circulation
The summary goes to the associate running diligence. Nothing reaches the wider deal team until it is approved.
Exactly what this agent can see, touch and change
The same five controls sit behind every ollo agent. These are this one's settings — visible before you build it, not buried in an admin screen afterwards.
Context
What reaches the modelDeal teams running many reference calls under time pressure, where the risk is that a dissenting voice on call three never makes it into the memo.
Escalation
When it asks a personAn associate approves the signal summary before it is circulated to the deal team
Applications & Rights
Which tools it uses, and what it may do in each
NotionRead only
DropboxRead only
Google DriveReadWrites newCreates new records or documents. Never edits, moves or deletes anything that was already there.Verification
How you know it’s rightEvery claim links to the document it came from. A statement the agent cannot cite does not make it into the output — which is what makes the result reviewable in minutes rather than re-read end to end.
Who it’s for
Deal teams running many reference calls under time pressure, where the risk is that a dissenting voice on call three never makes it into the memo.
What you’ll need
- Transcript folder connected
- Expert engagement policy agreed
- Diligence question written down
- Named approver on the deal
What you get
- A ranked shortlist of people to approach
- A cited aggregate of call signal
- A list of unresolved disagreements
What it doesn’t do
It does not arrange or run the calls, and it does not confirm that an expert is cleared to speak to you. Compliance clearance stays with your team, and the agent only works from transcripts you give it.
We build the first one with you
Not a template you configure alone. We sit with your team, build it on real data, and hand over the controls.
Scope
One session with the people who actually do the work. We agree what the agent reads, what it may write, and who approves.
Co-build
Built on your own data, not a sandbox. You watch it being made, so you know why it behaves the way it does.
Handover
You own the controls. Change the context, tighten the rights, move the approval gate — without coming back to us.
Parts of the work currently spread across the categories below. It does not replace any of those products outright.
Frequently asked
How does it decide who to rank first?
By closeness to the diligence question you wrote in step one: the role held, how recently, and whether their experience touches the specific claim you are testing. The ranking comes with the reason attached, so an associate can disagree with it in a glance.
Do transcripts get stored somewhere new?
No. They stay in the folder your team already uses and the agent reads them there. The aggregate it writes goes back into the same deal folder, under the same permissions.
What stops it from over-reading one strong opinion?
Every point in the aggregate is attributed to the call it came from, and a point made once is shown as a point made once. Where calls conflict, both versions appear side by side rather than being resolved for you.
Can it handle notes that were typed rather than recorded?
Yes. Handwritten-then-typed notes, partial transcripts and email follow-ups all go into the same folder and are read the same way. Quality of input shows in the output, so the agent marks where a point rests on a sparse note.