AI for consultants and accounting firms.
Kylon turns the discovery call into a proposal, the engagement into a readout and client documents into a close pack. Partners check every number before it leaves the firm.

One engagement, from the discovery call to the next phase.
From the use case Consulting Firms: From Discovery Call to Next Phase.
- 01
Turn the first call into a deal
Ruth Calder, COO of Pinecrest Freight, books a discovery call on Cal.com. Orin reads the Read AI report: dock scheduling runs on phone calls and spreadsheets at 3 sites, and Ruth wants a 6 week diagnostic. Orin adds the $36,000 deal to the Pipeline tab.
Who decidesLeah reads the notes and tells Orin to draft the proposal.

- 02
Draft the proposal
Sam points Orin to the scope in the call notes. Orin drafts the proposal in PandaDoc: interviews at 3 sites for $9,000, a process map for $12,000, a workshop and 90 day roadmap for $15,000. $36,000 fixed, half at signing.
Who decidesLeah sends the proposal, and Ruth signs it on Friday, Feb 5.

- 03
Plan the work and write the charter
Orin builds the Asana workplan, 6 weeks and 14 tasks with the readout on Fri, Mar 12, and drafts the charter. Leah takes out the promised 30% cut in dock wait and $180,000 a year in savings. Orin saves her note as an engagement rule: never promise savings we have not measured, say what we will measure.
Who decidesLeah approves the charter before it is shared with Ruth.

- 04
Sort the workshop output
Sam drops the board export into the thread. Orin sorts 38 stickies into 4 clusters with dot votes, with dock bookings by phone on top at 9 votes, and pulls 7 action items with owners and due dates from the Read AI notes.
Who decidesSam owns the workshop output and hands the dock logs to Mei.

- 05
Write the readout
Orin maps the process in 4 lanes with 3 pain points and sets the baseline from Mei's dock log analysis: 47 minute average dock wait, 11 touches per load, 82% of 1,240 loads booked by phone, 312 overtime hours a month. It drafts the readout email from Leah's Gmail with the engagement rule applied.
Who decidesLeah sends the readout email to Ruth.

- 06
Close out hours and invoice
Mei asks Orin to close out the hours. Orin pulls the project from Harvest: 164 of 180 budgeted hours used, 16 left, $18,000 uninvoiced. It drafts invoice NA-1042, 2 of 2, for $18,000. NA-1041 was paid Feb 12.
Who decidesLeah sends every invoice.

- 07
Propose the next phase
With Phase 1 closed and $36,000 paid, Orin drafts Phase 2 in PandaDoc: an 8 week pilot at Ridgefield for $48,000, with a table of what we will measure against the February baseline and no promised savings.
Who decidesLeah reviews the draft and walks Ruth through it on Mon, Apr 5.

What Kylon runs for professional services
2 use cases, each with the steps, the tools and a set-up prompt.
The tools you already use.
Each use case here connects some of these. Anything with an API can be connected the same way.
Gmail
Cal.com
Read AI
PandaDoc
Asana
Harvest
QuickBooks
Xero
Karbon
DocuSign
Dropbox
Outlook
Google Drive
Common questions3
Can AI draft a consulting proposal?
Yes. The agent drafts scope, phases and fees from the discovery call, and a partner edits and sends it.
How does it help an accounting firm?
It files what clients send against each engagement, drafts the chase for what is missing and assembles the close pack. The partner checks every number and signs it.
Does client data stay inside the firm?
Work happens in your own Kylon workspace with the access you grant to each agent, and nothing leaves the firm without a person sending it.
Other jobs, other teams.
Start with one workflow this week.
Open a Room, connect the tools above and paste the set-up prompt from any use case on this page. Kylon builds it with you one step at a time.

