ActivePieces Automation (Notion + Slack) – Human-in-the-Loop Workflows

We are looking for someone with proven, hands-on experience using ActivePieces to build a production-ready automation system connecting Notion and Slack.

The goal is to keep our Notion project database clean, ensure no deadlines are missed, and allow team members to resolve issues directly from Slack using natural-language replies.

This project must be built entirely in ActivePieces.
Do not apply if you intend to use Make.com, n8n, or any other automation platform.

We will not provide example workflows or step-by-step documentation.
You should already know how to build this.

Our Setup (High Level)
Notion
- One “Global Project Database”
- Key properties include:
- Name (Title)
- Priority (Low, Medium, High, URGENT)
- Status (OPEN, WORKING, TESTING, CLOSED)
- Date (due date)
- meta_client (client name or “Internal”)
- Assigned To
- Created by
- Type, Department, Created Time

Slack
- One Slack channel per client
- Slack users correspond to team members in Notion
(user names may not always match exactly between systems)

What Needs to Be Built
You may choose the internal architecture, as long as the system reliably does the following:

1. Scheduled Monitoring
- Poll the Notion database on a configurable schedule
- Detect:
- Overdue items (Date is in the past AND Status is not CLOSED)
- Items missing required fields (Priority, Status, Date, meta_client, etc.)

2. Overdue Item Slack Alerts + Human-in-the-Loop Updates
- Alert the correct Slack channel based on meta_client
- @mention the appropriate user(s) responsible for the item
- Allow natural-language replies, including:
“It’s actually done”
“I need 3 more days”
“Push this to Friday”

The system must:
- Capture replies
- Parse intent
- Update the relevant Notion fields (Status, Date, etc.)

If a reply cannot be confidently interpreted, the system should request clarification rather than failing silently.

3. Missing Metadata Slack DM Flow
- Send a direct Slack message to the “Created by” user
- Clearly identify which fields are missing
- Accept natural-language free-text replies
- Update the correct Notion fields based on the response

4. Slack Reply Handling, Execution, Timeouts & Logging

This flow should handle all downstream processing, including:
- Listening for Slack replies
- Correlating replies to the correct Notion item
- Resolving Slack users to Notion users, accounting for cases where names do not exactly match
This may be handled via email matching, stored mappings, or intelligent/fuzzy matching logic
- Parsing intent (status updates, date changes, metadata updates, etc.)
- Executing updates in Notion
- Handling 24-hour timeouts if no reply is received
- Logging all activity to an ActivePieces Table, including:
- Notification sent
- Human response received
- Action taken
- Timeout

The system must safely support multiple simultaneous active notifications without replies being misrouted.


Implementation Expectations
- ActivePieces only
- Native Notion and Slack pieces where appropriate
- JavaScript and HTTP steps are encouraged for:
- User correlation logic
- Text parsing
- Advanced handling
- OpenAI steps (Pieces) may be used if helpful (preferred LLM)
- Must be production-ready, not a proof of concept

Deliverables (Required)
The final delivery must include four separate ActivePieces workflow exports (JSON files):
1) Scheduled Notion Monitoring & Issue Detection
2) Overdue Item Slack Alerts + Human-in-the-Loop Resolution
3) Missing Metadata Slack DM Flow
4) Slack Reply Handling, Notion Updates, Timeouts & Audit Logging

Also required:
- A short setup guide (Loom video or document) showing credential setup and field mapping
- One brief handoff call or chat for final questions

How This Will Be Tested
We will:
1) Create a fake overdue task
2) Create a task missing required fields
3) Reply to Slack messages in natural language (including status and date changes)

If the correct Notion fields update and all events are logged correctly, the project is complete.

Before You Apply
- Do not apply if you plan to use another automation platform
- We will not provide example workflows or instructions
- You should already be comfortable building this in ActivePieces

In your application, briefly include:
- Your experience using ActivePieces in production
- One example of a human-in-the-loop automation you’ve built