Understand your work OS

Your CRM, informed by conversations. Less manual entry.

A customer calls, clarifies a request on WhatsApp and waits for a reply. AlefOS connects the received exchanges to the relationship, so an authorized colleague can find the need, the promise and the next step. Your AI CRM draws on captured work, with information your team can review.

AI CRM: an internal use, not your public catalogue.

This page explains how authorised colleagues retrieve a relationship, its exchanges and next steps. This internal scope is separate from the initial customer application, which presents published offers and collects enquiries.

Workflow example

The call ends. The customer story continues.

  1. Camille asks whether the quote includes maintenance. The connected call brings the available exchange into the company environment.

  2. The customer context retains the question and the promised written reply. Your team reviews the details rather than starting from a blank note.

  3. Alex takes over and asks the compatible assistant what remains open. He checks the terms and prepares the reply within his authorized scope.

Fictional illustration. Available capabilities, sources and access depend on your environment; conditions are explained below.

On this page
  1. A relationship is more than a set of fields
  2. Find the right person
  3. Understand what remains open
  4. Prepare the next step in the relationship
  5. Start with a question, not a list
  6. What reduces CRM data entry?
  7. Tools and prerequisites

A relationship is more than a set of fields

A customer expresses a budget, preference, time constraint or objection during an exchange. Searching fields remains useful, but does not necessarily recover that story.

AI CRM describes access to AlefOS relationship memory through an assistant. It uses information that is actually available and authorized, alongside contact details and structured data.

Find the right person

“Who was looking for a two-bedroom apartment near the centre?” The search should find the context in which that need was expressed, then check that it is still relevant. An old preference does not automatically become a current need.

Contact searches can use a name, phone number, email address or identifier. Signal searches cover recorded context. Results do not provide access to the company’s entire database.

Understand what remains open

Before following up, review the latest useful interaction, outstanding documents, objections, products discussed and open missions. The contact profile brings these elements together when they are connected and accessible.

“Who is waiting for a reply?” and “Why is this file blocked?” lead to different actions. The source and date of the context help distinguish them.

Prepare the next step in the relationship

A response can be drafted from customer context and validated business information. Preparation is not sending. The recipient, content and channel must match the action that is actually authorized.

AI CRM does not automatically replace specialist software or its regulatory fields. It makes operational context available for consultation and use in supported AlefOS workflows.

Start with a question, not a list

Ask which expressed needs match a new offer, which objections remain unresolved, or which follow-ups need preparing. Contacts provides the authorized relationship context; AI CRM describes how to work with it. Review relevance, source and current circumstances before acting.

What reduces CRM data entry?

Start with the work that already happens: an authorized call, a received message or a meeting. The configured connections bring that material into the company environment. Useful information is linked to the customer, so the next person does not start with an empty record.

Your team still checks ambiguous information, corrects mistakes and adds what was never received. Updating a separate CRM requires a defined integration and field mapping; connecting an assistant alone does not populate every field in every CRM.

Operational reference

Modules, tools and access conditions

This reference describes MCP operations and their effects. The list exposed to your assistant depends on the deployed server, enabled modules and permissions. Reading, preparing and executing are distinct operations.

Contacts and relationship context

Search identity fields, read linked context and inspect confirmed products. Adding or enriching a contact is a write operation; the preview mode depends on the tool policy.

ToolFunction and effect
contacts.searchSearch contacts by identity fields, with cursor pagination and source filtering.
contacts.getRead one contact profile and its available relationship context.
contacts.search_productsFind confirmed products in the contact’s product memory.
contacts.draft_addAdd or enrich a contact; preview behavior follows the selected mode.

Read this module’s documentation →

Signals and controlled actions

Adding a signal can enrich memory and prepare pending work without executing an outbound action. An action_id identifies prepared work; its execution and result must remain distinguishable from a read-only answer.

ToolFunction and effect
signals.searchSearch extracted signal titles and content by keyword.
signals.getRead an accessible signal, its participants and linked missions.
signals.addEnrich memory from a structured note or source; missing details can require clarification. No outbound action is executed.
signals.updateCorrect or enrich an existing signal without creating another one.
alefos.briefingRead a compact briefing across visible actionable work and recent context.
alefos.daily_summaryRead a summary for one local calendar day within the current scope.
actions.confirmExecute previously prepared work identified by action_id after confirmation.

Read this module’s documentation →

Missions and responsibility

Read the source and execution state before acting. Confirming executes a pending mission; rejecting and archiving have different effects. Assignment follows company roles.

ToolFunction and effect
missions.listList the current user’s missions, optionally grouped by contact.
missions.getRead a mission’s source, related objects and execution state.
missions.confirmConfirm and execute a pending mission.
missions.rejectReject a pending mission without executing it.
missions.archiveArchive a mission without executing or deleting its action.
missions.assignAssign from the company queue or reassign with the required role; changes are audited.

Read this module’s documentation →

The scenarios on this page illustrate workflows, not real customer data or guaranteed outcomes. Connections and capabilities must be enabled and verified for your environment. Official business systems retain their responsibilities.

Continue exploring

A capability of your company’s environment. Your team can use this capability from a compatible assistant through its configured MCP access. The deployment defines the enabled tools and permissions. Understand your AI connection ↗

Explore one customer’s evolving relationship.

Compare ways to connect ChatGPT to your CRM ↗