Live Workflow Interviewer

The description states that Live Workflow Interviewer is a web-based tool built with TypeScript and React, deployed via OpenAI Sites. It uses:

• GPT-Realtime-2.1 for low-latency voice conversation over WebRTC;
• GPT-5.6 Luna to update the workflow graph through deterministic Graph Patches;
• GPT-5.6 Sol to audit readiness and produce an AI-assisted after-view.

The tool enables users to describe business processes in English or Japanese while looking at a shared workflow canvas. Each turn of conversation becomes part of an evolving, evidence-linked workflow. It includes:

• Natural, interruptible voice interaction;
• Real-time updates to the workflow graph via validated patches;
• Separation between “how work happens today” and “how AI might help”;
• An optional Codex-ready implementation brief and economic impact estimate based on user assumptions.

It is not tied to a specific domain like supplier invoice processing but includes that as a demo preset. No authentication or database was added; session data remains in the browser unless explicitly exported.

Inference This product appears to be a prototype or proof-of-concept for a new kind of workflow documentation and AI-assisted process design tool, not yet a commercial offering.

The author claims that Live Workflow Interviewer addresses a gap in how small businesses introduce AI — not because they lack documents, but because real work involves exceptions, judgment calls, and approval boundaries that are rarely written down. Traditional interviews separate conversation from documentation, often missing valuable tacit knowledge.

This positioning suggests a shift from static process mapping to live, collaborative, evidence-driven workflow design, where AI supports human judgment rather than replacing it.

Inference There is no indication of prior market positioning or branding beyond this self-description. The project appears to be in early-stage development with no commercial traction or customer feedback yet.

The description states that the tool targets small businesses who want to introduce AI but do not know where it can safely help. These users are described as having real work processes that contain exceptions, judgment calls, and approval boundaries — which are rarely documented.

It also mentions that the tool is designed for use by consultants or business operators, though no explicit ICP segmentation is provided beyond this.

Inference The target customer segment is likely small to mid-sized enterprises (SMEs) in process-heavy industries, such as finance, operations, or supply chain. However, there is no evidence of actual customers or personas defined.

There is no evidence of a business model or pricing structure in the description. The author does not mention any monetization strategy, subscription tiers, or sales channels.

The project is described as a hackathon submission and includes no indication of revenue streams, licensing models, or commercial partnerships.

Inference No business model has been developed or tested at this stage.

The author reports that the application:

• Is built with TypeScript and React;
• Uses OpenAI Sites for deployment;
• Implements WebRTC for real-time voice interaction;
• Relies on OpenAI API keys, including GPT-Realtime-2.1, GPT-5.6 Luna, and GPT-5.6 Sol;
• Includes structured model contracts, runtime validators, and fail-closed readiness gates;
• Supports bilingual (English/Japanese) interfaces;
• Offers offline export capabilities in HTML and Markdown;
• Uses per-runtime request throttling and project-budget limits.

It also includes features like:

• Evidence provenance checks;
• Stale-response protection;
• Collision-free workflow layout;
• Text fallback for microphone-unavailable environments.

Inference The technical stack is consistent with a developer-focused prototype or MVP. The system is designed to be lightweight, secure, and auditable, but lacks persistence or integration features.

There is no evidence of traction, customers, or adoption beyond the author’s own development and testing. The project is described as a hackathon submission with no mention of:

• Users;
• Revenue;
• Customers;
• Product usage metrics;
• Feedback loops;
• Market validation.

The author notes that future versions may include deliberate persistence, collaboration, and integrations — but these are not implemented in the current version.

Inference This is a pre-product or prototype stage project with no demonstrated traction or maturity.

There is no evidence of competitive analysis or awareness of existing tools in this space. The author does not reference other process mapping, workflow automation, or AI-assisted interview platforms.

The tool appears to be unique in its approach of embedding correction into the live interview loop and separating AS-IS from AI-assisted workflows.

Inference No competitive landscape is known or described. The product may be addressing an underserved niche, but this has not been validated.

• Unproven market fit: No evidence of customer feedback or real-world testing.
• Limited scope: The tool does not include persistence, collaboration, or integrations — features likely needed for commercial viability.
• No business model: No indication of how the product will generate revenue.
• Single-person team: The project is built by one individual (the author), which may limit scalability and development speed.
• Highly technical prototype: Not designed for general use, but rather as a proof-of-concept.
• Dependency on OpenAI APIs: Reliance on proprietary models and services introduces risk of changes or access limitations.

Inference The tool is in an early-stage prototype phase with significant risks around commercialization, scalability, and market validation.

1. What specific feedback have you received from small business operators or consultants who tested this tool?
2. How do you plan to validate that the live correction loop reduces follow-up interviews compared to traditional methods?
3. Have you identified any potential use cases beyond supplier invoice processing?
4. What are your plans for monetization and go-to-market strategy?
5. Are there any technical dependencies or risks related to OpenAI API availability or pricing?
6. How do you intend to scale beyond a single-person development team?