North

What the company appears to be: North is a self-reported project that claims to build an "intelligence layer" for companies to maintain coherence as they evolve. It connects to existing tools (documents, code, meetings, etc.) and organizes knowledge into three layers: Record, Direction, and Living Documents. The system allegedly detects changes, traces their impact across artifacts, and proposes grounded updates.

What changed: The project was submitted as part of the OpenAI 2026 hackathon. It is described as a proof-of-concept with a multi-agent development process using GPT-5.6 and Codex. No commercial traction or revenue data are provided.

Single most important open question: Is there evidence that North's core functionality—detecting meaningful change, tracing impact, and proposing updates—is technically feasible at scale, or is it an unproven concept?

Note: This analysis is based solely on the self-reported description provided by the authors. No external verification, revenue data, customer feedback, or product usage metrics are available.

The description states that North is an "intelligence layer" that keeps a company coherent as it evolves. It connects to tools like documents, presentations, meetings, email, calendars, and code. It organizes project knowledge into three layers:

• Record: what happened (meetings, emails, research, agreements).
• Direction: current decisions, assumptions, priorities.
• Living Documents: decks, plans, websites, product specs, financial models, code.

When something changes, North determines what changed, which artifacts are affected, and why they may now be outdated. It prepares grounded, reviewable updates while keeping the user in control.

It is not described as a change-detection tool alone but as a "shared context layer" that other people or agents can act on.

Claim: North is an intelligence layer for maintaining coherence across evolving company artifacts.

Evidence: Author's own write-up.

Inference: The system uses AI to extract decisions and build dependency graphs between sources, direction, and living documents.

The project positions itself as a solution to the problem of "project drift"—where teams spend hours updating information in multiple places because changes don’t propagate naturally. It claims to make projects feel “alive” by maintaining a shared understanding of what happened, what the team believes, and how each piece of work relates to that direction.

The authors state they were inspired by their own startup-building experience, where coordination problems became more severe as teams became leaner and AI enabled faster iteration.

They describe North not just as a summarizer or Q&A tool but as one that creates an evolving model of a project, understands how decisions connect to real work, and turns changes into concrete actions.

Claim: North solves the problem of organizational knowledge fragmentation.

Evidence: Author's own write-up.

Inference: The positioning implies North is a platform for managing shared context across tools and workflows.

The description does not name specific customers or personas. However, it suggests that North targets teams building startups or products where rapid change is common, such as those using lean processes and AI-driven development.

It appears aimed at organizations with multiple artifacts (e.g., docs, code, decks) that need to stay aligned when direction shifts.

Claim: North serves teams working in fast-moving environments where alignment across tools and documents is critical.

Evidence: Author's own write-up.

Inference: The target ICP likely includes product managers, engineering leads, startup founders, or project coordinators who manage evolving documentation and strategy.

There is no mention of pricing, monetization, or business model in the description. The authors do not describe how North would be sold, whether it’s a SaaS offering, or if there are any plans for commercial deployment.

Claim: No evidence of business model or pricing.

Evidence: Author's own write-up.

Inference: If this is intended to become a product, it may follow a SaaS model, but that is not stated.

North was built using GPT-5.6 and Codex throughout both the design and development process. It uses multi-agent workflows:

• A builder agent implements phases.
• An independent reviewer inspects code changes.
• A fresh agent tests the experience without prior knowledge.

It also builds an evidence-backed dependency graph between sources, direction, and living documents. The system evaluates whether new information represents a meaningful change, determines what existing direction it affects, identifies related artifacts, and proposes actions.

Claim: North uses AI to analyze connected project info, extract decisions, and build a dependency graph.

Evidence: Author's own write-up.

Inference: The use of multi-agent systems suggests an experimental or prototype approach rather than a production-ready solution.

There is no evidence of traction, revenue, customers, or adoption. The project was submitted to a hackathon and described as a proof-of-concept. No data on usage, retention, or product-market fit are provided.

Claim: No traction or maturity signals.

Evidence: Author's own write-up.

Inference: This is likely an early-stage prototype, not yet in production use.

The description does not mention competitors. It does not reference existing tools like Notion, Confluence, Airtable, or GitHub for managing project knowledge or alignment.

Claim: No competitive landscape described.

Evidence: Author's own write-up.

Inference: North may compete with tools that help manage documentation and collaboration, but no direct comparison is made.

• Unproven technical feasibility: The system relies heavily on AI to detect meaning in unstructured data (e.g., meeting transcripts), which may be unreliable at scale.
• Lack of real-world validation: No evidence of actual use cases or feedback from users.
• Unclear scalability: The approach of scanning every artifact with an AI model could be slow and expensive.
• Over-reliance on AI for decision-making: The system proposes actions but requires human confirmation, which may limit adoption if not well-integrated into workflows.
• No commercialization plan: No roadmap or monetization strategy is evident.

Claim: Technical and commercial risks due to lack of validation and scalability concerns.

Evidence: Author's own write-up.

Inference: These are potential barriers to product-market fit or adoption.

1. What specific types of artifacts does North currently support, and how does it handle different formats (e.g., PDFs, markdown, code)?
2. How does North distinguish between a confirmed decision and a discussion point in a meeting transcript?
3. Can you demonstrate how North traces impact from one change to multiple documents? What is the accuracy rate of these traces?
4. How does North manage conflicts when two artifacts contradict each other?
5. Are there any known limitations or blind spots in how North detects meaningful changes?
6. What are the current integrations, and what tools does it connect with today?
7. What is the plan for scaling this system beyond a hackathon prototype?

Claim: These questions aim to probe technical depth, scalability, and product maturity.

Evidence: Author's own write-up.

Inference: These are critical to assess whether North can move from concept to viable product.