Assistants and briefs
An AI step that ends in a grade and an action, as in Sales OS.
We have no published build on the OpenAI API. Our AI product Sales OS calls a Claude-compatible API and reports cost and tokens per workspace, and OpenAI keeps retiring model snapshots.
An AI step that ends in a grade and an action, as in Sales OS.
A model name kept in settings so a retirement is a config change.
Provider cost and token count reported per workspace.
Every score carries the note that produced it.
Illustration. A dashed outline marks where the technology would sit. We have no published build on it.
gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro snapshots.
GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini snapshots.
whisper-1 and the GPT-4o transcribe models.
gpt-5.3-codex, gpt-5.1 and gpt-5.4-nano.
Source: OpenAI API deprecations page, checked 4 November 2026. The page lists a recommended replacement for each model.

An engine that crawls a site, enriches it and writes a graded brief through a Claude-compatible API, with cost and tokens reported per workspace.

A chatbot project delivered at a fixed price, with GraphQL in the stack, beside Node.js and .NET back ends.
Python
React
Docker
Node.js
TypeScript
AWSThe deprecations page lists GPT-5 and o3 snapshots for 11 December 2026, GPT-4o and GPT-4.1 snapshots for 14 January 2027, transcription models for 26 February 2027, and gpt-5.3-codex, gpt-5.1 and gpt-5.4-nano for 1 April 2027.
No. We say so on the first call and show the nearest work, which is Sales OS on a Claude-compatible API and a fixed-price chatbot project.
In Sales OS the model sits behind settings and provider cost and tokens are reported per workspace, so a swap changes configuration and shows its cost.
You do, from the first commit. Repositories and cloud accounts stay in your name.
Thirty minutes on your data, your model and your costs. No sales pitch.