ThreadLine

• 

  Landing Page

• 

  Graph of truth

• 

  Check for contradictions

• 

  Ledger

• 

  See how the project drifts

• 

Inspiration

Teams already communicate everything needed to execute a project where: decisions happen in meetings, deadlines are mentioned in Slack, and ownership is assigned in emails. Yet this information is rarely a single, reliable source of truth.

Critical details are forgotten, and teams unknowingly operate on outdated plans. A deadline might change in one conversation. Traditional project tools depend on manual updates, and meeting summaries often stop at notes without actually aligning to them.

This gap inspired Threadline.

Threadline is an AI-powered coordination platform that transforms scattered conversations into a continuously updated map of project truth capable of capturing decisions, commitments, owners, deadlines, dependencies, risks, and contradictions.

What it does

Threadline allows users to create projects and import communication from sources such as: Slack-style messages Meeting transcripts Email threads Project briefs Stakeholder updates

The platform analyzes these inputs and extracts structured project data, including decisions, deliverables, owners, and deadlines.

This information is visualized in an interactive Project Truth Map, showing how all parts of a project connect. Instead of static task lists, users see relationships what depends on what, who owns each piece, and where risks exist.

Every extracted insight is linked to its original source, including quotes, timestamps, and confidence levels. It clearly distinguishes between confirmed facts, inferred insights, unresolved gaps, and contradictions, ensuring transparency and trust.

How I built it

Threadline is a full-stack web application built using Next.js, React, TypeScript, Tailwind CSS, and React Flow. The interface centers around a dynamic graph visualization powered by React Flow, enabling users to explore relationships between project elements through zooming, panning, and node interaction.

The frontend leverages reusable components, Framer Motion animations, and shadcn/ui for a polished user experience. The backend uses SQLite with Prisma to manage structured project data.

At the core is an AI extraction layer that converts unstructured conversations into structured JSON. Each extracted entity includes metadata such as source references, timestamps, confidence scores, and supporting quotes.

Challenges I ran into

One major challenge was translating natural language into an evolving project state. While extracting tasks is straightforward, determining whether new information updates, or contradicts previous data required careful modeling. I solved this by treating every change as a time-based event rather than overwriting existing data.

Visualizing complex project relationships without overwhelming users was also difficult. I addressed this through thoughtful graph design, distinct node types, and layered interaction controls.

Accomplishments that I'm proud of

I successfully built a system that transforms messy, real-world communication into a structured, interactive project model.

The Project Truth Map provides a clear, visual representation of how work connects, while features like contradiction detection and impact simulation offer insights that traditional tools lack.

I'm proud of the transparency built into the system where every insight is traceable back to its source, helping users trust and verify the AI’s output.

What we learned

I learned that the biggest challenge in project execution is not communication but shared understanding. Teams can have access to the same information yet interpret it differently.

I also learned that trust in AI systems depends on transparency. Providing evidence and clear distinctions between confirmed and inferred data is essential.

What's next for ThreadLine

Next, I plan to integrate Threadline with tools teams already use, including Slack, Gmail, Microsoft Teams, Jira, and Asana. I aim to enable real-time updates, automated alerts, and continuous synchronization between conversations and project state.

Built With

• css
• data
• engineering
• flow
• framer
• graph
• javascript
• motion
• node.js
• prisma
• prompt
• react
• rest
• shadcn/ui
• sqlite
• structured
• tailwind
• typescript