Promise Rader

Company: Promise Rader

Self-reported basis: The description is entirely self-reported and unverified, based on an author's own submission to the OpenAI 2026 hackathon on Devpost. No external corroboration or historical data is available.

What it appears to be: A prototype web application that processes plain text (messages, notes, emails) to extract “invisible commitments” and suggest next actions. It includes both a fast deterministic engine and AI-backed modes for extraction.

What changed: The project is a demo-level prototype built as part of a hackathon submission. No evidence of prior versions or evolution beyond this single iteration.

Single most important open question: Is there a viable commercial use case for extracting open loops from conversation signals, and if so, what would be the minimum viable product (MVP) that could gain traction with users?

The description states that Promise Rader is a React web app with a Node server/MCP layer, designed to process text and extract “invisible commitments” or “open loops.” It offers three extraction modes:

1. Fast logic: Uses offline, explainable heuristics (sentence/clause splitting, regex phrase families).
2. AI key mode (BYOK): Integrates with a user-provided chat-completions endpoint.
3. Local AI mode: Uses a WebGPU WebLLM model in the browser.

The app allows users to paste or upload text and returns:

• Categories/tags for each signal (e.g., “you promised”, “they asked”).
• Confidence scores.
• Evidence phrases from the input.
• Suggested next actions.

Inference: The product is a proof-of-concept tool, not a production-ready service. It is described as a prototype with no evidence of revenue, customers or adoption.

The author states that Promise Rader was built on the idea that “a lot of real work gets created as conversation signals” and that it aims to extract an “invisible backlog” from ordinary text.

It positions itself as a tool for:

• Extracting commitments, intentions, requests, decisions, and reminders.
• Turning these into actionable next steps.

The project is described as a demo-level prototype, not a commercial offering. The author does not claim any traction or product-market fit beyond the hackathon submission.

Inference: The positioning is exploratory and early-stage. It reflects an idea about task extraction from conversation, but no evidence of prior market validation or customer feedback exists.

The description states that the tool is intended for users who:

• Deal with everyday messages, notes, emails.
• Want to extract actionable commitments from these conversations.

It does not specify a particular industry or persona beyond “users” or “people in conversation.”

Inference: The ICP is not clearly defined. It appears to be aimed at individuals or teams managing informal communication workflows, but no evidence of customer segmentation or personas exists.

The description does not mention:

• Any pricing model.
• Revenue streams.
• Monetization strategy.
• Subscription tiers or usage-based billing.

It is described as a demo prototype, not a commercial product.

Inference: No business model or pricing evidence is present. The tool is not monetized, and no indication of how it would be sold or used at scale exists.

The project is built with:

• Frontend: React + Vite.
• Backend: Node.js server with MCP (Model Control Protocol) layer.
• AI integration:
  • BYOK mode via user-provided API key.
  • Local AI using WebGPU and WebLLM (Qwen2.5 model).
• Fast engine: Deterministic logic using regex, sentence splitting, and scoring signals.

It includes:

• src/App.jsx for UX orchestration.
• src/promiseEngine.js for fast extraction.
• src/byokAI.js and src/localAI.js for AI modes.
• Server-side tools like analyze_text and plan_open_loops.

Inference: The technical stack is functional for a prototype but not scalable or production-ready. No evidence of infrastructure, security, or scalability considerations.

The project is described as:

• A hackathon submission.
• A demo-level prototype.
• Not yet integrated into any workflow or product.

There is no evidence of:

• Customers.
• Revenue.
• Adoption.
• Usage metrics.
• Product iteration history.

Inference: No traction or maturity signals are evident. The project is at a very early stage, with no indication of user engagement or commercial viability.

The description does not mention:

• Competitors.
• Existing tools in the same space.
• Market positioning relative to others.

It is described as a new idea, not a product that competes with existing solutions.

Inference: No competitive context is provided. The project appears to be exploratory and unanchored in an existing market or competitive landscape.

1. No commercial viability evidence: The tool is a prototype, not a product.
2. Unproven use case: No evidence of demand for extracting invisible commitments from conversation.
3. Limited scope: Only one team member (Sohan Poudel) is involved.
4. No monetization path: No pricing or business model described.
5. Technical limitations: AI modes depend on user-provided keys or browser support; fast engine may lack nuance.
6. Unverified claims: The author’s own description is unverified and self-reported.

Inference: The project is high-risk, exploratory, and not yet proven to have a viable product-market fit or commercial application.

1. What specific user pain points are you trying to solve with this tool?
2. Have you tested the prototype with real users? If so, what feedback did you get?
3. How do you plan to validate demand for extracting open loops from conversation?
4. What is your roadmap beyond this prototype?
5. Are there any existing tools or workflows that this would integrate with?
6. What are the privacy and data handling implications of processing user messages?