The Bionic Sales Rep: Human + AI Workflow Diagram
How a sales rep and AI share the work: a workflow diagram of the whole sales process, what AI can do alone, what needs the rep's yes, and where to start.
Table of contents
What is a Bionic Sales Rep?
A Bionic Sales Rep is a salesperson who lets AI take the part of the job that is not selling: research before the meeting, notes after the meeting, CRM fields and a first draft of the follow-up. The salesperson keeps what requires a human: listening, judging, negotiating and building trust. How much time that frees up depends on what your salespeople's week looks like today, so the first step is to measure it.
The term is our working label, not a standard. The idea behind it is old: let the machine take the routine so the human can take what requires judgement. What is new is that AI can today read and write ordinary Danish, and so it can take over tasks that used to require a person.
TL;DR:
- Sales reps spend a small part of the week on the actual selling. The rest goes to CRM, research, internal meetings and follow-up.
- It makes sense to let AI prepare, write up meeting notes and suggest updates, while the rep decides and sends.
- Start with one task, build it properly, measure before and after with your own numbers, and then expand.
Why do sales reps spend so little of the week selling?
Research from, among others, Salesforce (State of Sales) has repeatedly shown that sales reps spend less than half of their working week talking to customers and selling. The rest disappears into tasks everyone knows: updating the CRM, preparing meetings, looking for information, writing follow-ups and taking part in internal meetings.
It is not the salespeople's fault. Most sales organisations have built up their tools one at a time, and between the tools sits a person copying. If you want to see what this costs you, you can work it out without any model: take a salesperson's total salary cost and multiply it by the share of the week that does not go to customers. That figure is not what AI can save, but it shows how much of the salary today goes to something other than selling.
The number has to be measured at your company. Ask your salespeople to write down, for one week, what they spend their time on, in broad categories. It takes a few minutes a day, and it gives a starting point that no one can dispute afterwards.
How do the sales rep and AI share the work?
The simplest way to see it is to go through a sales process step by step and ask two things: what may the AI do on its own, and what requires the sales rep's yes? That division is also the one we use when we build solutions for customers: you decide what the system may do by itself. Everything else waits for an approval.
Workflow: who does what in a bionic sales process
| Phase | The sales rep's role | What the AI may do on its own | Requires the sales rep's yes |
|---|---|---|---|
| Research before first contact | Assesses whether the company is a fit, and whether the timing is right | Gather publicly available information and put it together on one page | The decision to make contact |
| Meeting preparation | Reads the overview and adjusts their angle | Pull earlier emails, notes and agreements from the CRM and put them together | Anything that changes an agreement |
| During the meeting | Listens, asks, negotiates | Take notes (if everyone is informed and has said yes) | Nothing, here the rep is alone |
| Right after the meeting | Reads through and corrects | Write up the notes and suggest next steps | The notes, before they are shared |
| CRM update | Approves | Update fields according to your own rules | Changes to price, stage and probability |
| Follow-up | Adjusts tone and content, sends themselves | Write a first draft based on what the customer said | Anything that goes out to the customer |
| Pipeline review | Decides priorities | Collect open offers and flag those without a new reply | What should be done about them |
The last column is the decisive one. Nothing that goes to a customer is sent without a human's yes. That is both a reassurance for the salesperson and a necessity, because AI can make mistakes, especially when it has to reproduce figures, names and agreements.
The four parts of a bionic working day
A bionic setup is not one tool. It is four tasks that can be solved separately, and each of which can be the first step.
Preparation: the sales rep does not turn up unprepared
Preparation is often what gets cut when the calendar is full. Here AI can gather what already exists: earlier correspondence, notes from earlier meetings and public information about the company. The result is a short page, not a report. The salesperson should be able to read it in a few minutes and judge for themselves what is relevant. Anything the AI finds on the internet must have a source, so the salesperson can see where it comes from.
The conversation: the sales rep is present
During the meeting the sales rep should listen, not write. If you choose to let AI take notes, all participants must be informed about it, and you must have control over the rules for recording. We go through this in our guide to call recording in Denmark. Without recording, the salesperson can instead dictate a short status after the meeting, which the AI turns into notes.
The follow-up work: notes, fields and next steps
This is where most salespeople lose time, and it is also where AI does the most good. The notes are written, the CRM fields are proposed, and the next follow-up is put in the calendar. The salesperson reviews, corrects and approves. The goal is not that nobody has to read the notes, but that the sale is recorded without the salesperson having to spend the evening on it.
Learning: better conversations over time
When conversations are written out (and you have permission to do so), you can learn from them. Which questions do the best reps ask? Which objections keep coming back? This is valuable for new salespeople, who can hear and read how experienced colleagues handle a difficult situation. But it requires caution. If the analysis feels like surveillance, you lose the salespeople's trust, and then nothing works. Be open about what is analysed, who sees it and what it is used for. Under the EU AI Act, it is also, as a rule, prohibited to use AI to infer employees' emotions from voice or face in the workplace (with narrow exceptions), so keep the analysis to the content of the conversations.
What is required before it works?
Three things determine whether a bionic working day holds.
The CRM is already in use. If your salespeople do not use the CRM consistently today, AI will not solve the discipline problem. It will just automate a gap. Start by agreeing what each phase of the pipeline means and which fields must be filled in.
There is one place where the truth lives. If agreements are in emails, spreadsheets and the CRM at the same time, the AI does not know which version applies. Decide what the source is.
The salespeople are involved. The best wording is the one that is true: the AI proposes, you decide. The sales rep does not lose control, but is spared some of the tedious work. If it feels like control from outside, the tool will not be used.
What does it look like in practice?
We have built a sales cockpit for our own sales that shows the principle on a small scale: a call list for the salesperson. The conversation is written out in Danish, the AI proposes the outcome (for example "Call again") and writes a draft email based on what the customer said. The salesperson calls, the AI writes the transcript and the email draft, and the salesperson sends it themselves.
On the use cases page you can see more examples of what we build. The most relevant here is "The meeting is over. The CRM work begins": the agent listens in on the salesperson's own sales meetings, writes the notes, updates the pipeline in the CRM and puts the next follow-up in the calendar. It may write the notes and update the fields itself, but it asks first before anything goes out to the customer. It can be connected to HubSpot, Salesforce, Pipedrive and Microsoft Dynamics, as well as telephony and calendars.
The examples on the page are illustrations with invented names and figures. They show how a solution is built, not what it will deliver for you.
If an AI is also to make the calls itself, that is a separate track. Here we work with JesperAI, and the rules for automated calls must be assessed legally before anything is set in motion.
How do you get started?
There is no reason to begin with a large overhaul. Five steps go a long way.
- Measure your starting point. Have the salespeople log their time for one week. Also measure how long it takes from the end of a meeting until the CRM is updated.
- Choose one task. The one where the time spent is greatest and where the rules are clear. For many, it is the notes and CRM after the meeting.
- Build it finished before you build further. A task that works every day is worth more than four half-finished ones.
- Measure again. Compare with the starting point from step 1. Also ask the salespeople whether they use it.
- Expand. Take the next task when the first one runs without supervision.
What does it cost?
There is no standard price, because the tasks are different. The price of getting a solution built depends on how many systems have to be connected, how much approval logic has to go in, and how much of your data has to be cleaned up first. We give an estimate after a short conversation about the task. If you would rather learn to build it yourself, there is a personal AI mentoring programme, where you work on your own tasks.
FAQ: Questions about the Bionic Sales Rep
Q: What is the difference between a Bionic Sales Rep and a salesperson who uses AI tools? A salesperson who uses a single AI tool now and then gets an isolated gain. In a bionic setup the tools are connected to your CRM and to each other, so the notes end up in the right fields and the follow-up is put in the calendar without anyone copying anything. The difference is the connection, not the number of tools.
Q: Do sales reps lose control when AI takes on part of the work? No, if you build it correctly. The AI prepares, writes drafts and proposes. The salesperson decides, corrects and sends. Nothing that goes to a customer should be sent without a human's yes.
Q: How do you ensure GDPR compliance when conversations are recorded and analysed? You need a valid basis for recording, you must inform the people involved before the recording, store only what is necessary and have a retention period. If the data is processed by a supplier, there must be a data processing agreement. We go through the rules in our guide to call recording in Denmark. If you are in doubt, ask a lawyer who knows your industry.
Q: How long does it take to get a bionic working day to work? It depends on the task and on how tidy your systems are. A single, delimited task such as notes and CRM after meetings is quicker than an overall overhaul of the entire sales process. That is why we start with one task and build it finished before we take the next.
Q: Does it only work in B2B, or also in telesales to private customers? The principle applies in both places, but the gain lies in different places. With many short calls, it is the follow-up work that eats the time. With few, large deals, it is the preparation. Telesales to private customers is also subject to stricter rules on marketing and recording, so have that assessed first.
Q: Can it be connected to our current CRM? We build on top of the systems you have, including HubSpot, Salesforce, Pipedrive and Microsoft Dynamics. You do not have to change CRM to get started. If your CRM is not in use already, however, you should sort that out first.
Q: What does it cost? The price depends on the task, and we give an estimate after a short conversation. There is no list of fixed prices.
Conclusion: A bionic working day is about how the work is divided
The question is not whether sales reps should work with AI, but which parts of the work should move from a person to a machine, and which should never do so. What requires trust, judgement and empathy stays with the sales rep. What is routine can be moved. Start by measuring what your salespeople spend their time on, choose the task that takes up the most of it, and build it finished.
Book 20 minutes of free advice if you want to hear whether your sales task is worth building. We will be honest with you, including if it is not. See also examples of tasks we can help with.
See also:
- Call recording in Denmark: rules and GDPR
- What does an AI sales rep cost?
- How do you train your AI sales rep?
Sources
- Salesforce, State of Sales
- The Danish Data Protection Agency (Datatilsynet), guidance on recording telephone calls
- EU Regulation 2024/1689 (AI Act), Article 5
#ai-sales#workflow#productivity#augmented-sales#conversation-intelligence
Contents
Should your sales reps spend their time talking?
We have built it for our own sales: the call is transcribed, the AI suggests the outcome, and the email is ready. We look at how your sales run today.
Book a reviewShould your sales reps spend their time talking?
We have built it for our own sales: the call is transcribed, the AI suggests the outcome, and the email is ready. We look at how your sales run today.
How we built it · Sales & follow-up
Sales cockpit with AI
A call list for salespeople. The call is transcribed in Danish, AI suggests the outcome (e.g. “Call again”) and drafts the email based on what the customer said.
- The salesperson calls
- AI writes transcript and email
- The salesperson sends
Related articles

Own or rent software? A decision guide for Danish businesses
Own or rent software? Cost over time, data lock-in, GDPR and exit, self-host and fair source, plus a concrete decision framework for Danish businesses.

Claude Code in a business: what an AI developer agent can do, and what it requires
What Claude Code, Anthropic's agentic developer tool, can do in a business, and what it requires in security, governance and organisation.

AI strategy for businesses: a step-by-step guide
A practical guide to an AI strategy: map workflows, choose use cases, fix your data, run a pilot, bring staff along and know when not to automate.