Pay-per-message AI chatbot that simulates an interview with Garrett Berg.
What it does
This x402-enabled endpoint lets callers send chat messages to an AI-powered interview chatbot representing Garrett Berg, a technical operator focused on backend, product, and due diligence in crypto/infra. Each message costs up to 0.01 USDC (10,000 units of the 6-decimal USDC token) on the Base network, paid via the x402 exact payment scheme.
The endpoint accepts POST requests and returns a JSON object containing a session token, an expiration timestamp, and a success flag. This suggests a session-based conversation model where the initial paid call establishes or continues a chat session. The chatbot appears to be an alternative to booking a live call — the provider's personal site offers both a Calendly link and a link to "Conduct an AI interview with me."
Documentation is minimal: there is no OpenAPI spec, no dedicated docs page, and no detailed description of the request body format or conversation flow. The x402 challenge is live and well-formed, confirming the endpoint is operational. Callers should expect to experiment with the request payload (likely a JSON body containing a message or prompt field) since no input schema is published.
Capabilities
Use cases
- —Conducting an automated AI-driven interview with Garrett Berg before scheduling a live call
- —Agents autonomously screening or pre-qualifying a consultant via conversational AI
- —Demonstrating x402 micropayment-gated conversational endpoints
Fit
Best for
- —Recruiters or collaborators who want to learn about Garrett Berg's background asynchronously
- —Developers exploring x402 payment-gated chat APIs
- —AI agents that need to gather information about a person via structured conversation
Not for
- —General-purpose LLM chat — this is scoped to one person's interview persona
- —High-volume bulk querying — per-message payment adds up and session model is unclear
- —Users needing detailed API documentation before integrating
Quick start
curl -X POST https://www.garrett-berg.com/api/interview/chat \
-H "Content-Type: application/json" \
-H "X-PAYMENT: <x402_payment_header>" \
-d '{"message": "Tell me about your experience with backend infrastructure."}'Example
Request
{
"message": "Tell me about your experience with backend infrastructure."
}Response
{
"success": true,
"expiresAt": 1720000000,
"sessionToken": "abc123-session-token"
}Endpoint
Quality
The x402 challenge is live and well-formed with a clear output schema, but there is no published input schema, no OpenAPI spec, no docs page, and no pricing page. The request body format must be guessed. The example request JSON is inferred, not documented.
Warnings
- —No input schema published — request body format is unknown and must be inferred
- —No API documentation, OpenAPI spec, or usage guide available
- —Example request JSON is speculative based on the chatbot context
- —Session/conversation flow (how to continue a multi-turn chat) is undocumented
Citations
- —Garrett Berg describes himself as a technical operator in backend, product, and due diligence in crypto/infrahttps://www.garrett-berg.com
- —The site offers a link to 'Conduct an AI interview with me'https://www.garrett-berg.com
- —The x402 challenge requires up to 10000 units of USDC (0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913) on Base networkhttps://www.garrett-berg.com/api/interview/chat
- —The output schema returns success (boolean), sessionToken (string), and expiresAt (number)https://www.garrett-berg.com/api/interview/chat
- —Payment is sent to address 0xB013a7f5F82bEA73c682fe6BFFB23715bb58e656https://www.garrett-berg.com/api/interview/chat