1What it does
Each new prospect in Notion is scored against the agency's ideal client profile. Fits go to a private Telegram group for a human decision; the team's verdict goes back in as feedback.
2Who it is for
- Industry
- Agency
- Platform
- Make
- Category
- Lead follow-up
- Tools used
- MakeNotionTelegramOpenAI
3How it works
Every step, in order, is in Workflow at a glance, pinned on the right as you scroll.
- 01 Trigger
New record in the Notion prospects database.
- 02 Action
Assemble the prompt with Array Aggregator.
- 03 AI / action
OpenAI 'Create a completion' with GPT-4o-mini.
- 04 Action
Fit: send summary to Telegram for review.
- 05 Action
No fit: webhook restarts with the next record.
- 06 Action
Team marks qualified or not via the bot; feedback updates the ICP.
- Trigger
- New record in the Notion prospects database.
- AI step
- Evaluate the company against the ICP and return fit or no-fit with reasons.
4How it went
The author says it saves over 20 hours each month.
Only what the source reports. Numbers above come from the source.
5Worth knowing
- GPT-4o-mini was chosen to keep cost low across hundreds of records.
- A human makes the final call.
6Recipe
Written by Builds That Work The source did not publish this. We wrote it as a starting point.
Starter: "Here is our ideal client profile: {{icp}}. Here is a company: {{company_record}}. Return JSON {fit: true|false, score: 0-100, reasons: [..3 bullets]}."7Source
Open the original Blog case study by Ilya Bolkhovsky →
Source opened and read on 2026-10-11.