Turn scattered sales context into a decision a seller can inspect.
This native Mac application was built for sales work that arrived in several forms and at different stages. Instead of asking one model to produce a confident answer, the system separates research, qualification and drafting so the seller can see what is known, what is missing and what still needs judgment.
The problem was larger than a single prompt.
Useful account context arrived in email, documents and copied notes.
Staff rebuilt the opportunity record before deciding whether the account fit.
Missing information could be hidden by a polished summary or premature email draft.
The seller needed to keep control of qualification and external communication.
A useful system had to fit the work and its limits.
- Allow the user to edit the intake before an opportunity enters the queue.
- Keep research evidence attached to the account rather than collapsing it into unsupported prose.
- Compare evidence with configurable services and qualification criteria.
- Prepare stage-appropriate follow-up without sending an email automatically.
Each agent has a separate job.
Research, qualification and drafting agents share one evidence record. They can prepare a recommendation and a relevant email, but the seller decides whether the opportunity fits and what leaves the company.
Research agent
Finds public company context, records the source and leaves unavailable information unresolved rather than guessing.
Qualification agent
Tests the evidence against configurable services, criteria and disqualifiers. It separates a fit signal from a missing answer.
Drafting agent
Uses the evidence and sales stage to prepare an editable reply, question or follow-up. The seller decides what leaves the company.
Opportunity record
Keeps the intake, evidence, gaps, risks and drafts together so the reasoning can be reviewed later.
One traceable path from input to review.
- 01Import an email, document or clipboard note
- 02Edit the intake before creating an opportunity
- 03Compare the evidence with configurable services
- 04Review the gaps, risks and next email
Technology chosen around the operating environment.
Native SwiftUI and Swift application for macOS.
OpenAI Responses API with web search for sourced public research.
Local document, email and clipboard capture with an editable review step.
macOS Keychain for protected local access to application secrets.
Local persistence designed around the opportunity and its evidence record.
Evidence for the next decision, not a promised result.
Time from incoming context to an opportunity that is ready for review.
How quickly a seller can prepare a relevant next step after reviewing the evidence.
Important questions and fit criteria that remain unresolved.
Changes made to research, fit judgments and drafts before the opportunity moves forward.
This page describes an internal application and the way its agents divide the work. It does not publish private account information, sales performance or a claimed conversion improvement.
Start with the work, the constraint and the decision owner.
We can map the process, decide which agents would help and define a first release that is small enough to evaluate.
Discuss the problem