Web forms
Requests that land with the right contact
Sign-ups, questions and downloads come in as a lead or contact. If the contact already exists, your team gets a suggestion to update it instead of a duplicate.
for SMEs, non-profits and federations that work in a CRM
Your team still retypes requests and emails into Teamleader, HubSpot or Salesforce. We build an AI workflow that prepares the data and recognises existing contacts, and your team approves everything in a pilot of 4 to 6 weeks.
No obligation · 30 minutes · online, with someone who builds, tests and fine-tunes it
Ethic is new. The team behind it isn’t. Ethic is the AI studio of Zenjoy, a digital product studio from Leuven. Since 2011, Zenjoy has built and maintained software and integrations for more than 400 organisations, with Salesforce, SAP, Odoo and Airtable among others.
At a glance
Honest advice
Every major CRM now builds in its own AI. We’ll tell you honestly which of the three fits, even if that means you don’t need us.
| The AI in your CRM | An automation in your CRM | Your own workflow | |
|---|---|---|---|
| Good for | AI help inside your CRM itself. Focus AI in Teamleader drafts quotes and invoices and updates deals, Breeze in HubSpot answers customer questions, and Agentforce Nonprofit helps with participants and prospect research, among other things. | Fixed forms and fixed rules, without AI. A web form becomes a lead, a new status sets up the next step. | Free text, emails and PDFs that start outside your CRM, data from several systems, or a small European model. |
| Watch out | What you can use depends on your plan, and you often pay per use or in credits. Data that starts outside your CRM, such as a PDF in the mailbox, has to get in first. | Works with fixed fields. Whatever someone types in a single free-text box stays one block of text that your team has to untangle itself. | More building up front, and someone has to maintain it. That’s why maintenance is included. |
| What we do | Checking which data your CRM’s AI may use, making agreements and training your team. | Help set up forms and rules. If that’s enough, we don’t build AI. | Build with an AI fact sheet, test with your real requests and maintain after launch. |
What you can do with it
Web forms
Sign-ups, questions and downloads come in as a lead or contact. If the contact already exists, your team gets a suggestion to update it instead of a duplicate.
Emails
An email gets a label, is linked to the right contact and gets a suggested reply. Your team reads it through and sends it themselves.
Reports
The figures come straight from your CRM. AI writes a first explanation, with a reference for each figure to the report it comes from.
Data quality
AI looks for contacts that are probably the same person, or where fields are missing. A colleague decides what gets merged.
How it works
This is how the integration for requests works during the pilot.
Through the form on your website, an email or a PDF attachment.
The AI extracts the name, organisation, question and campaign from the text, even if someone typed everything in one box.
By email address, name and organisation. Found? Then a suggestion to update the contact follows. Not found? Then a suggestion for a new lead.
In the pilot, your team approves every suggestion. Two possible contacts, a missing field or an unusual question get a clear flag. Merging is always done by a person.
For each request, we keep the original text, what changed and who approved it.
In the pilot, we measure two things: how much time a request costs your team, and how often a colleague had to correct a suggestion. With those figures, you decide whether to continue, and whether clear-cut cases can then go into your CRM on their own. Filling in a field or creating a lead can be undone. Merging contacts and emails to the requester stay with your team.
AI fact sheet
Every integration comes with an AI fact sheet. In five minutes, your DPO and your IT lead can read what happens to your data.

Example
Website requests to Teamleader
What it does
Reads requests from the website form and the info@ mailbox, fills in fields and looks for an existing contact or company. Suggests an update or a new deal. Sends no emails to the requester itself and does not merge contacts itself.
What data
Name, email address, company, the question and how the requester found you. No quotes, no invoices, no special categories of personal data.
Model and location
Teamleader remains the source system. For each request, only the fields above go to a language model that runs in the EU. Which model is recorded in the AI fact sheet.
Who checks
During the pilot, the sales lead approves every suggestion. After that, clear-cut cases may go through on their own if the pilot figures allow it. Merging and anything that goes to the requester stays with your team.
What we keep
For each request: the original text, the suggested fields, the decision and who approved it. That log is kept separately from Teamleader, in a database in the EU. Retention period: 90 days.
Owner and stopping
Your organisation owns the code, the prompts and the keys. Revoke the API access in Teamleader and the integration can no longer read or write anything.
Control
The integration gets its own user or API key in your CRM, with only the data it needs. You can revoke that access yourself.
We test with duplicate contacts, half-completed forms, spam and a CRM that briefly doesn’t respond. CRM packages limit how many API calls you can make, and we take that into account.
Only the fields from the AI fact sheet go to the model. We keep the log with the original text, suggestion and decision separately, with a fixed retention period.
Where your CRM runs is set out in your contract with the vendor. HubSpot and Salesforce offer a European data location. Our own part runs in the EU by default.
Who builds it
Ethic is Zenjoy’s AI studio, based in Leuven. The same team builds and maintains your workflow, and everything is in your name.
15 years
building and maintaining software
400+
organisations
4 years
using AI daily in our own work
After the pilot
01
With every major update of your CRM, we test the integration again.
02
If a better or more efficient model comes along, we switch without rebuilding the workflow. We do test it again.
03
Code, prompts, documentation and access are in your name.
04
Updates, monitoring and small changes are covered by a monthly agreement. New workflows are planned separately.
After a first conversation, you get a proposal with a clear scope and a fixed price.
Frequently asked questions
Question not listed? Just ask us.
A data processing agreement and security overview are available on request.
After the introductory call, we choose one task together, for example the requests from your website form. We build a pilot of 4 to 6 weeks around it. Your team works with it in your CRM straight away and approves every suggestion. At the end, you see how much time a request still takes and how often a suggestion had to be corrected. With those figures, you decide whether to continue.
Your data stays in your own CRM. Where it runs is set out in your contract with the vendor. HubSpot and Salesforce offer a European data location; at Salesforce it is called the Hyperforce EU Operating Zone. The AI part we build runs in the EU by default. Provider and country are in the AI fact sheet, and the AI only gets the fields it needs. If you use your CRM’s own AI, your vendor’s terms apply. We review them together with your DPO.
Yes, and we take that into account. The AI suggests, a colleague decides. In the pilot, your team approves every suggestion, and merging contacts is always done by a person. Only after the pilot, with the figures in hand, do you decide whether clear-cut cases can go through without review. Everything is logged, so you can find errors and we can adjust the workflow.
It happens regularly. As part of maintenance, we test the integration with every major update of your CRM, and we get an alert if something goes wrong. Zenjoy has been maintaining platforms and integrations for years after launch; an API that changes is part of that.
That depends on what we build and under whose name it’s put into use. In the AI fact sheet, we set out who is the provider and who is the deployer. Anyone deploying AI must also take measures to support the AI literacy of the staff who work with it (Article 4). Training is the simplest step for that. Scores or selections of people are at most a suggestion from the AI: a person always decides. This isn’t legal advice: have it checked by your DPO or legal adviser.
No. Salesforce says the Nonprofit Success Pack will remain supported. For new customers, Salesforce recommends Nonprofit Cloud, now called Agentforce Nonprofit. We build on the data model you use today and record in the documentation which objects and fields the integration uses. If you switch later, you’ll see straight away what needs to change with it.
Often, yes. If you work with Odoo CRM, see our page about Odoo. For Pipedrive, Microsoft Dynamics or your own database, we check whether there is an API or a secure export, and tell you honestly whether an integration is feasible.
Related
Workflow
Forms and requestsSector
Non-profitsIntegration
Microsoft 365In the introductory call, we look at your CRM and one task your team retypes today. You get honest advice: your CRM’s AI, an automation or your own workflow.
No obligation · 30 minutes · online, with someone who builds, tests and fine-tunes it