DockPilot

Company: DockPilot

Self-reported basis: The analysis is based entirely on the project description provided by the caller — its name, tagline, the author's own write-up, and technology tags. No external verification or historical data are available.

What it appears to be: A read-only, evidence-first scheduling tool for dock and shipyard operations that integrates live public data, deterministic logic, and AI-assisted planning. It is not a system for writing to schedules or making safety decisions.

What changed: The project was submitted as part of the OpenAI 2026 hackathon. It represents an early-stage prototype focused on validating a controlled pilot with a commercial shipyard.

Single most important open question: Is there sufficient evidence that the target customer (shipyard operators) will adopt this tool, or is it still a proof-of-concept?

The description states that DockPilot is an evidence-first decision-support workspace for reviewing dock and shipyard scheduling changes. It brings together:

• Schedule data
• Facility facts
• Vessel-data gaps
• Tide context

Into a read-only workspace.

It allows operators to:

• Select an existing booking or define a new-contract scenario
• Describe the change
• Compare the published baseline with a candidate schedule

The system uses deterministic logic to preserve all published phases and detect shared-resource overlaps. It does not overwrite schedules or fill missing data with guesses.

It also integrates AI-assisted planning, using GPT-5.6 Terra via OpenAI-compatible API, but only for interpreting operator intent into constrained schedule adjustments — it cannot write to the official schedule or make safety decisions.

Not evidenced: No information on actual customers, revenue, pricing, or adoption.

The description states that DockPilot is not a vessel-movement authorization system, and that authorized operators retain control of the official schedule. It is designed to support review processes, not decision-making or execution.

It positions itself as:

• Evidence-first
• Deterministic in logic
• AI-assisted, but not autonomous
• Read-only workspace

The author emphasizes that it is not a system for writing to schedules or making safety decisions — these remain with authorized operators.

Inference: The positioning suggests a niche, compliance-oriented tool aimed at planners who need to validate changes against public data and constraints. It does not appear to be a general-purpose scheduling platform.

The description states that DockPilot is designed for shipyard operators, specifically those involved in reviewing dock scheduling changes.

It is intended to support:

• A controlled, read-only pilot with a commercial shipyard
• Validation of schedule changes against public data and constraints
• Use by planners and operations teams

It is not described as targeting end-users or general maritime stakeholders — it is focused on operators who already have access to official schedules.

Not evidenced: No information on whether the target customer has adopted this tool, or if there is a defined ICP beyond "shipyard operators".

The description does not state any business model or pricing structure. It only describes the tool’s functionality and technical architecture.

It mentions:

• A controlled pilot with one dock and narrow historical data contract
• Back-testing against completed and cancelled bookings
• Expansion after UAT by planner and operations teams

Not evidenced: No information on monetization, customer acquisition, pricing tiers, or revenue model.

The system is built using:

• Frontend: React + TypeScript
• Deployment: Netlify Functions
• Data sources: Public Canadian data (Esquimalt Graving Dock bookings, Canadian Hydrographic Service tide predictions, dock dimensions)
• Backend logic:
  • Server-side adapters for retrieving and validating public booking and tide data
  • Browser keeps only a verified public snapshot for offline use
  • Deterministic schedule engine with test coverage
• AI integration: GPT-5.6 Terra via OpenAI-compatible API, used only for interpreting operator intent into constrained adjustments
• Safety logic: No AI or deterministic code can write to the official schedule or make safety decisions

Not evidenced: No information on scalability, data ingestion pipelines, or long-term architecture.

The project is described as:

• A hackathon submission (OpenAI 2026)
• A controlled pilot with a commercial shipyard
• Designed to start with one dock and narrow historical data contract
• Back-tested against completed and cancelled bookings
• A read-only prototype, not yet in production

It is not described as having:

• Customers
• Revenue
• Live users
• Product-market fit

Not evidenced: No evidence of traction, adoption, or usage beyond the hackathon submission.

The description does not mention any competitors. It is not clear whether DockPilot is addressing a gap in existing dock scheduling tools or if it competes with systems that already exist for maritime operations.

It is described as:

• Not an authorization system
• Not a general-purpose scheduler
• Focused on evidence-based review

Not evidenced: No information on competitive landscape, existing solutions, or differentiation from other tools.

• No traction or revenue: The tool is described only as a hackathon submission and pilot-ready prototype.
• Limited scope: It is designed for one dock and narrow historical data contract — not scalable to broader use cases.
• AI dependency without autonomy: AI is used only for interpretation, not decision-making — this may limit its perceived value.
• No monetization strategy: No pricing or business model described.
• Unproven adoption: No evidence that target customers will adopt the tool beyond a pilot.

Inference: The project appears to be in early-stage development and is not yet proven in real-world use.

1. What specific shipyard or operator is involved in the planned pilot?
2. How does the team plan to scale from one dock to multiple docks or facilities?
3. What are the key assumptions about user behavior and adoption that need validation?
4. Is there a defined path to monetization or customer acquisition beyond the pilot?
5. How is data governance handled for non-public sources, if any?
6. What are the main technical limitations of the current deterministic engine?