FlowLens Drift

The description states that FlowLens Drift is a one-page application built with Next.js and TypeScript. It compares two process descriptions:

• Written process: the intended procedure
• Real work: what people actually do

It identifies drift between these, links findings to source lines and quotations, separates supported evidence from uncertainty, prioritizes operational gaps, asks one high-leverage follow-up question, allows humans to add context, and reassesses while keeping added information visibly unverified.

The system uses the OpenAI Responses API with GPT-5.6 Terra, Zod schema validation, and structured outputs for analysis. It includes input validation, error handling, and a demo video capture system.

Evidence The author's own write-up.

Inference This is an MVP built for a hackathon; it does not appear to be a production-ready product or platform with customers.

The description states that FlowLens Drift was created to make visible the gap between written procedures and real-world operations, without pretending AI already knows the full truth. It is positioned as a tool for identifying process drift, verifying evidence, and asking follow-up questions rather than making automated decisions.

It emphasizes:

• Evidence-grounded loop
• Separation of checked evidence from uncertainty
• Human judgment remains central
• Not to evaluate or blame individuals

Evidence The author's own write-up.

Inference The positioning is rooted in operational improvement, not AI automation or decision-making. It reflects a shift toward transparency and human-in-the-loop systems.

The description does not name specific customer segments or personas. However, it implies that the target users are those working with process documentation and real-world operations — such as compliance officers, process analysts, or operational managers who need to identify discrepancies between intended workflows and actual execution.

It also suggests a focus on organizations looking for continuous improvement rather than one-time audits or evaluations.

Evidence The author's own write-up.

Inference Based on the problem described (process drift), the ICP likely includes internal process teams, compliance departments, or operational improvement units within enterprises. No explicit customer names or use cases are provided.

There is no evidence of a business model or pricing structure in the description. The project is presented as an MVP built for a hackathon and does not mention any monetization strategy, subscription plans, or sales channels.

Evidence Not evidenced.

Inference If this evolves into a product, it may be sold to enterprise clients via SaaS or consulting models, but no such indication exists in the current description.

The system is built using:

• Next.js (server-side API route)
• TypeScript
• React
• OpenAI Responses API with GPT-5.6 Terra
• Zod for schema validation
• Vitest for testing
• Playwright for demo capture and visual checks
• Codex used for scaffolding and refinement

It includes:

• Source-line and quotation verification
• Input-length and empty-input validation
• Safe handling of authentication, rate-limiting, timeouts, parsing, and error conditions
• Human-controlled follow-up submission
• Clear separation between verified and unverified context

Evidence The author's own write-up.

Inference The technical stack indicates a modern web application with strong emphasis on data integrity and user control. It is not a plug-and-play solution but rather a prototype designed for demonstration and iteration.

There is no evidence of traction, revenue, or customer adoption beyond the author’s own demonstration. The project was submitted to a hackathon (OpenAI 2026), and the MVP is described as a one-page application with 54 automated tests and linting checks.

No mention of users, usage metrics, or product-market fit.

Evidence Not evidenced.

Inference This is an early-stage prototype, likely not yet in production or used by any organization beyond its creator.

The description does not reference competitors or similar tools. It does not state whether there are existing solutions for comparing written procedures with real-world work or for identifying operational drift.

Evidence Not evidenced.

Inference The space may include process mining, compliance monitoring, workflow automation, and operational analytics platforms. However, no competitive positioning is evident in the description.

• No traction or revenue: The project appears to be a hackathon MVP with no evidence of real-world usage.
• Unverified claims: All descriptions are self-reported and unverified; there is no third-party validation.
• Limited scope: The current MVP only compares one written process and one description of real work.
• No monetization strategy: No indication of how the product would be sold or funded.
• Single-person team: A solo developer may limit scalability, iteration speed, or long-term maintenance.

Evidence Self-reported, unverified.

1. What is the intended user journey beyond the MVP?
2. How does the system handle edge cases or ambiguous inputs?
3. Has there been any feedback from potential users or domain experts?
4. Is there a plan to scale beyond the current one-page prototype?
5. What are the key assumptions about process drift and evidence collection that underpin this tool?
6. Are there any plans for integrating with existing enterprise systems (e.g., ERP, workflow tools)?
7. How is data privacy and security handled in the MVP?