AI workflows in your CRM, process first

    We audit your sales process, then put a fixed rule where a rule is enough and an AI step where the work needs judgement.

    Price
    Audit €2,200
    Invoicing
    Per phase, in advance

    Make Certified PartnerMake Certified Partner

    For teams who recognise this

    • After every sales call someone writes up the notes and the proposal by hand.
    • Requests from the website arrive as email and get retyped into the CRM.
    • Your team already uses AI in separate chat windows, and nothing comes back into the CRM.
    • Automating lead follow-up has been on the list for a year, and nobody knows where to start.

    What you get

    • Process audit
    • From call to proposal
    • From inbox to lead
    • Workflows between systems
    • AI safely on your data

    What it delivered

    20%+

    more margin, through automation and a B2B funnel next to the B2C sales

    Leads respond 30% more often.

    Evans & Watson, premium tea and herbs

    What we build

    • Process audit

      We walk your process from request to invoice and write down, per step, who does what and where it waits.

    • From call to proposal

      With Surge Quote the transcript of a sales call becomes a proposal you steer with prompts and share as a link.

    • From inbox to lead

      Surge Forms and Surge Chat write a request straight into the CRM, with an owner and a follow-up task.

    • Workflows between systems

      Scenarios in Make.com that connect your CRM to email, calendar and accounting.

    • AI safely on your data

      Surge Connect links an AI model to your CRM, and you decide what the model may read.

    How a project runs

    1. Intake

      A conversation about your process and about who struggles with what.

    2. Audit, 7 to 10 days

      We map every step, count where work is retyped or waits, and write the findings down.

    3. Roadmap

      Per step we decide whether a rule, an AI step or a person does it, and in which order we build.

    4. Build and test

      We build in phases and test with real examples from your own CRM before anything goes live.

    Price

    What it costs

    Audit €2,200

    The audit takes 7 to 10 days and ends with findings and a roadmap. The build is priced per phase after that, and you pay each phase before it starts. Your own AI agent starts from €4,997, in 10 working days, with 30 days of support.

    To be honest

    When AI is still too early

    When your CRM is full of duplicate records or the stages do not match how you sell, AI mostly does what already went wrong, faster. Then we start with the data and the setup, and the AI comes later.

    What is the difference between a workflow and an AI step?

    A workflow follows a fixed rule: if this happens, do that. An AI step handles work where the input is text and the outcome needs judgement, such as summarising a call or reading what a customer actually asks.

    The question we ask per step is simple. Can you write the rule down in one sentence? Then it is a workflow, and a workflow is cheaper, faster and predictable. Do you need to read, interpret or write? Then an AI step can do the first draft, and a person checks it where a mistake costs money.

    Step in the processRule or judgementWhat we build
    A new lead gets an ownerRuleRound-robin or per region, as a workflow in the CRM
    A deal moves to the next stageRuleTasks and a follow-up email per stage
    A sales call becomes a proposalJudgementAI step in Surge Quote, a person reviews before it goes out
    A request through the website becomes a leadRule (Surge Forms) or judgement (Surge Chat qualifies)Surge Forms writes the form straight into the CRM, Surge Chat qualifies in the conversation
    A manager asks how the pipeline is doingJudgementA model with read access through Surge Connect
    An invoice goes out after signingRuleA Make.com scenario to your accounting

    Three examples from that table, in more detail.

    Transcript to proposal. Before: someone listens back to the call, types the notes and builds a proposal in a document. After: the transcript goes into Surge Quote, you steer the draft with prompts, and you share the proposal as a link. The rep still decides what is in it.

    From shared inbox to lead. Before: a request arrives by email in a shared inbox and someone retypes name, company and question into the CRM. After: the request no longer arrives by email. Surge Forms replaces the email address on the site with a form that writes straight into the CRM, with an owner and a task, and Surge Chat qualifies a visitor in the conversation. So the inbox is replaced, not read. Nobody retypes.

    A CRM question to a model. Before: a manager exports a list to a spreadsheet to answer one question. After: a model reads the CRM through Surge Connect, with a scope you set, and answers in plain language.

    What does a process audit give you?

    A process audit gives you a list of findings and a roadmap: what we would fix, in which order, and whether each step gets a rule, an AI step or stays with a person. It takes 7 to 10 days and costs €2,200.

    We look at the route from request to invoice. Where does a lead come in, who picks it up, what gets retyped, where does it wait, and what happens after the signature. We do that in your own CRM and your own inbox, not on a slide. A findings table looks like this. The rows below are an example, not client data.

    FindingWhat it causesProposalType
    Web requests arrive by emailRetyping and requests that sit in the inboxForm writes straight to the CRMRule
    No fixed owner for new leadsSome leads wait daysRound-robin with a taskRule
    Proposals built from scratchHours per proposal, different every timeProposal from the call transcriptAI step
    Deals stall without anyone noticingPipeline looks fuller than it isAlert after a set number of daysRule

    Not every finding needs building. Sometimes the answer is to stop doing a step. You keep the roadmap, also if you build it yourself or with someone else.

    Where does a person keep watching?

    A person checks wherever a mistake costs money, reaches a customer or cannot be undone. That is our rule of thumb, and we write it down per step in the roadmap.

    In practice that means a proposal from Surge Quote is reviewed before it goes out. An AI step that classifies a request writes to a field, and the workflow only acts on it when the classification is clear. A model that reads your CRM gets read access only to what it needs, through Surge Connect with OAuth 2.1 and a scope per workspace. And every automation that sends email has an exit condition, so it stops when the customer replies.

    • Draft by AI, decision by a person, for anything that reaches a customer.
    • Write AI output to a field first, so you can see it before a workflow uses it.
    • Limit what a model may read, per workspace.
    • Log what an automation did, so you can trace a mistake back.

    Which tools do you use?

    We build in the CRM you already use, mostly Pipedrive, and use Make.com for the steps between systems. Sales Surge is a Make.com Certified Partner and we build and run scenarios in 17 Make organisations.

    For the AI steps we use our own Surge products: Surge Quote for proposals, Surge Forms and Surge Chat for intake, and Surge Connect when a model needs to read CRM data. Where an integration needs more error handling than a scenario gives, we build middleware on our own infrastructure. More on the combination in our guide to Make and Pipedrive and in the sales automation knowledge base.

    In short

    AI workflows, in short

    AI workflows are steps in your sales process where software does the retyping, sorting or drafting, and a person checks what matters. We start with a process audit of 7 to 10 days for €2,200. That gives you a roadmap: per step a fixed rule, an AI step or a person. We then build it in your CRM, with Make.com where the CRM stops.

    Source: Sales Surge price list, checked on 16 November 2026. Prices exclude VAT.

    Frequently asked questions about AI workflows

    Where do you start with AI in your sales process?

    With the process, not the tool. We walk the route from inquiry to invoice and count where your team loses time retyping, searching and chasing. Only then do we decide per step whether it needs a fixed rule, an AI step or a person. An audit costs 2,200 EUR and takes 7 to 10 days.

    How is this different from an AI sales agent?

    An AI sales agent is one agent that prospects, qualifies and follows up. We build it in a 10 working day sprint, from 4,997 EUR. AI workflows go wider: they sit throughout your process, from a call transcript that becomes a proposal to a chat that qualifies a lead and puts it in your CRM.

    Can an AI model just read our CRM data?

    Not just like that. We connect a model through Surge Connect, an MCP server with OAuth 2.1 and a defined scope per workspace. So you decide what a model may read and what not, instead of the model seeing everything. If you want to see what that gives you first, we start with one workspace and a limited set of fields, and only widen it once you have judged the outcome.

    When is Make.com not enough?

    Make.com fits steps between systems with a clear trigger, such as an invoice after a signature. It gets difficult when no message may be lost, not even when the other system is down for a while, or when many messages come in at once. Then we build middleware on our own infrastructure, with an outbox that keeps failed messages and offers them again, and a failure notice to a named person. That costs more to build, so we only choose it when it is needed.

    Does this work without a CRM?

    No. We do not take on work without a CRM component, because an AI step with no place to store the outcome mostly produces loose text. The outcome of an AI step, such as a summary or a score, belongs in a field on the deal or the person. Then a workflow can react to it and you can check later what the model did. If you do not have a CRM yet, the project starts with the setup. A CRM implementation costs from 2,950 EUR and takes 2 to 4 weeks.

    How long until an AI workflow is live?

    The audit takes 7 to 10 days. After that it depends on the roadmap. Lead routing or a follow-up email per stage is quick to set up, while an AI step that drafts proposals needs more testing with your own examples. The lead time per phase is in the proposal, and you pay each phase before it starts. That way you see something working after every phase.

    What data goes to an AI model?

    Only what that one step needs. For a proposal that is the call transcript and the deal details, not your whole customer base. Through Surge Connect you set per workspace what a model may read. When a model writes something back, it lands in a field first, so someone can see it before a workflow acts on it.

    See what your team can stop doing by hand

    The free growth scan shows within 48 hours where your team retypes and keeps lists.