ClarityHub

ClarityHub helps teams protect their most valuable resource—attention. It cuts through message overload, ranks what matters, and shows the ‘why’ behind every task.

Comment

Inspiration

ClarityHub is inspired by the idea of how large teams get buried in Slack chaos almost all the times. Talented engineers bounced between @mentions, scrolled endlessly to find the one urgent request, and lost focus every few minutes. The pattern was clear: cognitive overload. We didn’t need another chatbot; we needed a space that made communication explainable, auditable, and human again.

———

What it does

ClarityHub turns Slack noise into a real-time attention flightdeck:

  • Priority Inbox ranks work using urgency, impact, and confidence scores, exposing evidence chips and clarifying questions for every task.
  • Attention Tax Report shows interrupt timelines, deep-work blocks, hot zone users, and metric notes that explain missing data.
  • Scenario Demo Controls let judges inject curated incidents (e.g., payment outages, exec requests) and watch the entire pipeline update live.
  • Task Detail surfaces full Slack context, related docs, and associated issues so managers can act quickly.

———

How I built it

  • Agents & AI: Gemini 1.5 Flash powers a multi-agent pipeline—noise filter, task extractor, supervisor, priority engine, evidence validator.
  • Search & Storage: Elastic Cloud stores raw Slack-like messages and derived tasks, combining vector embeddings with BM25 for hybrid search.
  • Analytics: A deterministic Python module computes attention tax, timelines, and deep-work metrics, adding notes whenever data is insufficient.
  • Cloud Architecture: FastAPI backend on Google Cloud Run (secrets in Secret Manager); React frontend on Firebase Hosting; demo scenarios injected through a protected /api/demo/seed.

———

Challenges I ran into

  • Modeling realistic Slack scenarios without leaking internal data.
  • Making analytics trustworthy—both mathematically and narratively.
  • Keeping pipeline output explainable so teams can audit every decision.
  • Designing a judge-friendly experience that demonstrates “live” behavior with a single click.

———

Accomplishments that I'm proud of

  • An end-to-end attention pipeline where every score is backed by evidence and timestamps.
  • A deterministic analytics engine that surfaces meaningful deep-work metrics and explains missing values.
  • A demo workflow where judges can inject scenarios, watch the Priority Inbox populate RED/YELLOW tasks, and see analytics spike in real time.
  • Clean deployment on Google Cloud (Cloud Run + Firebase Hosting) with Elastic and Gemini as core components.

———

What I learned

  • Explainable beats magical—teams trust automation when they see the rationale.
  • Attention metrics come alive when you tie them to human stories, not just numbers.
  • Slack data is messy; building deterministic analytics forced me to understand message structure, timestamps, and team behavior at a granular level.
  • Designing for judges requires the right balance of realism, control, and narrative.

———

What’s next for ClarityHub

  • Plug directly into live Slack workspaces with real-time ingestion.
  • Sync tasks with Jira or other ticketing systems for full-loop workflows.
  • Add adaptive analytics that learn individual focus patterns and recommend scheduling changes.
  • Package ClarityHub as a managed service so remote teams can reclaim their attention in minutes.

Built With

Share this project:

Updates

Submission history