What Sales as Code is

The idea

Most of a revenue team's week is recurring work that everyone agrees matters and nobody has time to do well: reading up on an account before a call, writing outbound that is not generic, reconciling a forecast, checking which renewals are drifting, answering the same security questionnaire again. It gets skipped, or it gets done badly at 11pm.

Sales as Code gives that work a home before there is a person for each job. It starts from your website: the first launch reads it, drafts a business brief you confirm, researches the market and the first companies and contacts, and leaves addressed drafts in your own mailbox for you to send. There is one role per job, nine of them, sharing that confirmed brief, and each runs on a published schedule with every run recorded.

You still own the offer, the customer conversations, the commercial decisions and every approval. When you hire, the person you hire is assigned the role and inherits the brief, the research and the conversations in progress. The product is a way to start and run the motion before a full team exists, not a claim that a team is unnecessary.

How it works

• It is hosted. The work runs on our servers. There is no app to install and no plugin to keep updated. You buy a seat, connect your tools, set the schedule, and read what came back.
• It is draft-first. Agents prepare drafts and stage them for approval. Nothing reaches a prospect, a customer, or a channel without an explicit yes from a person. Outbound-capable roles also carry volume governors, so an approved sequence cannot quietly wreck your deliverability.
• It is vendor-agnostic. You configure your own company in Settings, and the product rebrands its interface, its prompts, and its exports around you. Nothing assumes a particular CRM or a particular way of selling.
• It is programmable. The same tools the scheduled agents call are exposed over a public API, as REST and over the Model Context Protocol. See the API reference and the OpenAPI specification.

What we do with your data

Because the work runs on our servers, we process the data you connect: your CRM, your inbox, your calendar. Each token is stored encrypted and read only on the schedule you set. We do not sell that data, share it with advertisers, or use it to train models, and you can revoke any connection at any time.

The full picture, including our subprocessors and our compliance posture stated as it is today rather than as a roadmap, is on the trust and security page. The legal versions are in the privacy policy, the terms, and the DPA.

Where to go next

• Pricing: one seat, every role, what it costs.
• The workflow: from your website to the next buying conversation, in four stages.
• Watch it work: runs recorded in the open.
• Docs: setup, concepts, and a page per feature.
• Free GTM audit: a read on your motion before you buy anything.
• Contact: reach a person.
• llms.txt: the same story, written for an AI assistant.