Keeping decisions connected to delivery.
My team moves fast, handling new projects, improvements, fixes, and experiments in parallel. Requests, decisions, and dependencies accumulate across Slack, meetings, Asana, and Notion. Keeping those conversations aligned with actual work requires repeated context gathering and follow-up.
I’m responsible for the team’s delivery and technical decisions. I turned my own coordination practices into a tool that handles the routine follow-up.
How it works.
Each source has a dedicated agent that collects and groups relevant activity, including mentions, threads, reactions, task changes, meeting transcripts, and project documentation. A central orchestrator connects their findings, identifies related events, and decides what needs attention.
Slack agent
Threads, mentions, reactions
Asana agent
Tasks, comments, status changes
Meeting agent
Transcripts, decisions, commitments
Notion agent
Notes and project documentation
Orchestrator
Connect related events, identify missing follow-up, and decide the next action.
Asana coordination project
Create tasks, assign owners, update statuses, comment, and escalate.
Closing the gaps after a conversation.
A meeting commitment, an idea worth evaluating, or a newly discovered blocker can become an Asana task with context and source references. Unanswered requests get an owner and a deadline. Relevant discussions can be surfaced even when nobody directly mentions the engineering team.
The orchestrator also compares task status and deadlines with what is being said elsewhere. It can flag inconsistent updates, identify dependencies, suggest priorities, and follow up with developers or escalate to me. Actions happen in its Asana project, keeping coordination visible alongside the team’s work.
An illustrative example.
A sync establishes a release date. A Slack thread later mentions missing API access, while the related Asana task still appears on track. Together, these signals suggest a dependency that could delay delivery. The orchestrator can connect the discussion to the task, record the blocker with source links, and alert the responsible people.
More context for the decisions that matter.
The system reduces the routine effort of collecting updates, capturing commitments, and chasing unanswered requests. It gives the team a traceable record of what was discussed, who owns the next step, and where progress needs attention.
Although this case focuses on an AI engineering team, the system can be just as useful in any department when adapted to its domain and workflows.