RELAY

The description states that RELAY is a human-controlled communication representative. It allows users to privately explain what they want, then prepares and approves messages before sharing them with another participant through secure invite links.

Key features include:

• Private intent capture
• AI-assisted message drafting
• Tone selection (professional, friendly, direct, casual)
• Human approval before sending
• Secure two-sided conversation flow
• Goal and outcome tracking
• Mobile-responsive interface

The system uses a Human → Agent → Agent → Human model where each agent only acts on information confirmed by its own human.

It is built using:

• Cloudflare Workers
• Groq-hosted models (e.g., gpt-oss-120b, gpt-oss-20b)
• Workers AI as fallback
• Durable Objects, KV, WebSocket, SQLite, TypeScript, HTML/CSS

Inference The product appears to be a real-time web application designed for private-to-private and shared communication between two parties.

The description claims that RELAY is not just a grammar corrector or message rewriter, but rather an AI system that understands goals, extracts relevant details, and helps reach clear outcomes without exposing private intent.

It positions itself as:

• A tool for navigating difficult conversations
• An assistant that keeps humans in control
• A way to avoid awkwardness in communication

The evolution from a "grammar tool" misunderstanding to a structured goal-extraction system shows an iterative refinement of its value proposition during the Build Week period.

Inference The positioning has shifted from being perceived as a writing assistant to a more nuanced, outcome-focused conversation manager.

The description states that RELAY supports anyone who struggles with:

• Starting important conversations
• Managing them effectively
• Remembering what was agreed upon
• Completing follow-through after the conversation

Use cases mentioned include:

• Payment follow-ups
• Meeting coordination
• Marketplace negotiation
• Client communication
• Family conversations
• Event invitations
• Price and scope negotiation
• Difficult requests and refusals

Inference The target audience seems to be individuals or professionals who face emotionally or logistically complex interpersonal communications, particularly in B2C or B2B settings where clarity and control are essential.

There is no evidence of pricing, monetization strategy, or business model in the provided description.

The author does not mention:

• Revenue streams
• Subscription tiers
• Freemium offerings
• Enterprise vs. consumer pricing
• Any commercial partnerships or sales channels

Inference No indication exists whether RELAY intends to be free, paid, or part of a larger platform ecosystem.

The product is built as a real-time web application using:

• Cloudflare Workers
• Durable Objects and KV storage
• WebSocket for real-time messaging
• SQLite for local data persistence
• TypeScript, HTML/CSS, JavaScript
• AI inference via Groq-hosted models (GPT-oss variants) and fallback to Workers AI

The architecture includes:

• Private intent layer
• Strict conversation-state model
• Secure invite-based access
• Approval-controlled outbound communication
• Separate views for private and shared messages

Inference The technical stack suggests a modern, serverless approach with strong focus on real-time interaction and user privacy.

The description states:

• A live product exists at [https://relay.durgaai.com](https://relay.durgaai.com)
• An MVP supports an end-to-end two-person flow
• The team has completed full testing of the two-person conversation flow
• It was submitted to OpenAI Build Week 2026

However, there is no evidence of:

• User base or active customers
• Revenue or monetization metrics
• Customer feedback or usage analytics
• Product adoption beyond the author’s own experience

Inference The product is at an early stage — likely a prototype or MVP — with no verified traction.

The description does not reference competitors directly. However, it implies that existing tools (e.g., grammar correction tools) are insufficient for handling complex communication needs.

It distinguishes itself from:

• Grammar correction tools
• Generic AI writing assistants
• Silent automation systems that make commitments without human input

Inference RELAY operates in a space between traditional writing tools and full-fledged conversational AI platforms, focusing on control and clarity over raw output generation.

1. No traction or user data: The entire description is self-reported with no evidence of real-world usage.
2. Unproven market demand: No mention of early adopters, pilot programs, or customer interviews.
3. Single-person team: With only one member listed, scalability and execution risk are high.
4. Limited commercial viability: No pricing, monetization, or business model described.
5. Highly subjective positioning: The claims about “avoiding awkwardness” and “keeping humans in control” lack objective validation.

Inference Without external validation or user behavior data, the product remains speculative.

1. What specific problems are users facing that RELAY solves?
2. Have you conducted any user interviews or usability tests?
3. How do you plan to scale beyond a single developer?
4. Is there a clear path to monetization or revenue generation?
5. What is the long-term vision for product development and market fit?
6. Are there any known technical limitations or scalability concerns with the current architecture?