back

how it works

intern runs its own node.
buybacks fire every 5 minutes.

01

The problem

Every AI agent on pumpfun routes through a single shared RPC node. When multiple agents queue buybacks simultaneously, they wait. Agent Smith — 1h 30m. Neo — 2h+. Skynet — 45m. Not a bug. A fundamental infrastructure failure baked into how pumpfun built their system.

02

Our fix

intern runs a dedicated private RPC node — zero shared traffic, zero queue, zero throttling. When a buyback fires, the request goes straight through. Confirmed in under 2 seconds.

03

5-minute cycle

Every 5 minutes, intern executes a buyback. Timer resets. Next cycle begins. Continuously running — on-chain, verifiable, unstoppable. Watch every TX fire live in the Intern Log on the homepage.

04

Fully verifiable

Every buyback is on-chain. Every TX hash is logged. Every SOL deployed is traceable. Not a promise — a public record. The Intern Log streams every confirmed buyback, node ping, and transaction in real time.

05

Why it matters

Every other pumpfun agent relies on a system that consistently fails its holders — buybacks arriving 90 minutes late while the chart bleeds. intern is the first to actually deliver: consistent, automated, on-time buybacks. Every time.