Language es en

Let’s talk

The catalogue

The cases we implement.

The cases we have put into production most often, ordered by the kind of work they solve. Yours come from looking at how you work, and rarely match this list exactly.

Business and operations

Daily work with documents, customers and content.

Desk work and documents

What is prepared by hand today — documents, reports, admin — stops starting from scratch every time.

Client presentations Every proposal is adapted by hand from the previous client’s.

Your own template as the source, with an agent bringing that client’s data and your previous proposals.

  1. Step no longer done by hand: You open the last client’s proposal The template fills with that client’s data
  2. Step no longer done by hand: You change names, figures and references by hand Your previous proposals arrive as reference
  3. You reread it in case something slipped through You reread it and decide what stays in

Work per proposal, in this example

3 h 45 min

What a person still does

Deciding what is offered and at what price.

Who uses it
Sales and management
What is needed
The template in an editable format and access to past proposals.

Entry service: AI implementation

Reporting and follow-up The board report is rebuilt by hand every month and lands late.

The report is generated from queries against your data and comes out laid out in your own format.

  1. Step no longer done by hand: You export from two or three systems The data gathers itself from every system
  2. Step no longer done by hand: The sheet gets reconciled by hand The report comes out in the usual format
  3. It lands late and with last week’s data It is ready when it is meant to be read

Work per report, in this example

6 h 45 min

What a person still does

Reading the number and deciding.

Who uses it
Management and area leads
What is needed
Read access to the source systems.

Entry service: AI implementation

Minutes and project documents What gets decided in the meeting lives in the head of whoever took notes.

Minutes, plan and next steps written as the meeting ends, in the project’s own format.

  1. Step no longer done by hand: Someone takes notes while people talk The meeting comes out transcribed and ordered
  2. Step no longer done by hand: They write them up if there is time The minutes carry the project’s format
  3. Half of it never gets written down You reread and confirm what was agreed

Work per meeting, in this example

1 h 10 min

What a person still does

Confirming what was actually agreed.

Who uses it
Whoever manages projects and whoever takes part in them
What is needed
Somewhere decisions are stored and the format you use.

Entry service: AI implementation

Processing documents and invoices Every month the same filling in, reconciling and filing comes round again.

The monthly cycle automated on your own tools, with a human review before anything is signed off.

  1. Step no longer done by hand: Every document gets opened one by one The document arrives read and classified
  2. Step no longer done by hand: The data gets typed into the system The data goes into the system already recorded
  3. It gets reconciled against what was there A person validates before signing it off

Work per monthly close, in this example

12 h 3 h

What a person still does

Validating and sorting out what does not add up.

Who uses it
Admin and finance
What is needed
Access to the inbox or document manager and to the system of record.

Entry service: AI implementation

Data analysis without going through IT Every business question is a ticket to IT and three days of waiting.

An analytics assistant that answers and shows the calculation it got there with. AA · AI Analytics Assistant · See it on the platform

  1. Step no longer done by hand: A ticket gets opened with IT The question is asked in plain language
  2. Step no longer done by hand: You wait your turn in the queue The answer arrives with the calculation in view
  3. A spreadsheet arrives once the question has already changed You ask again until it makes sense

Work per question, in this example

2 h 5 min

What a person still does

Asking well and deciding on the answer.

Who uses it
Business, without going through IT
What is needed
An accessible data model and permissions defined by you.

Entry service: Custom development

Customers, sales and support

Every conversation arrives with the history in view, and stops depending on who picks it up.

The inbox Email takes the morning and what matters surfaces late.

An agent over the inbox: it sorts, prioritises and leaves the draft with the history in view.

  1. Step no longer done by hand: The inbox gets read top to bottom Mail arrives sorted and prioritised
  2. Step no longer done by hand: Each one gets judged urgent or not The draft is written with the history in view
  3. What matters surfaces mid-morning You reread, adjust and send

Work per day, in this example

2 h 40 min

What a person still does

Deciding what gets answered and how.

Who uses it
Whoever handles the incoming mail
What is needed
Access to the mailbox and a written idea of what counts as urgent for you.

Entry service: AI implementation

Preparing offers and quotes Answering a tender ties up two people for days.

Requirements pulled out of the tender and a proposal built with your pricing and scope criteria.

  1. Step no longer done by hand: The whole tender gets read for requirements Requirements come out pulled from the tender
  2. Step no longer done by hand: A similar proposal gets copied The proposal is built with your criteria
  3. Price and scope get adjusted by hand You check price and scope before sending

Work per tender, in this example

16 h 4 h

What a person still does

Deciding the price and the scope.

Who uses it
Sales and management
What is needed
Access to past offers and to your pricing and scope rules.

Entry service: AI implementation

Sales email Personalising every email does not scale; sending the same one to everyone does not work.

One-to-one drafting over your own list, with each contact’s state in plain view.

  1. Step no longer done by hand: Each one gets written by hand, or the same goes to everyone Each email is drafted for that contact
  2. Step no longer done by hand: Track is lost of who was written to The state of each one stays recorded
  3. Follow-up depends on remembering You decide who gets a second push

Work per campaign, in this example

6 h 1 h

What a person still does

Deciding who gets written to and when.

Who uses it
The sales team
What is needed
The list wherever it lives and the criteria for what gets said to whom.

Entry service: AI implementation

Customer and supplier support Every reply means rebuilding the customer’s history first.

A reply drafted with the history and the tone of the house, ready to approve. EA · Engagement Agent · See it on the platform

  1. Step no longer done by hand: The customer’s history gets looked up The history arrives with the question
  2. Step no longer done by hand: Someone checks what was said last time The reply comes out drafted in your tone
  3. The reply gets written from scratch A person approves it and it goes

Work per reply, in this example

25 min 5 min

What a person still does

Approving and adjusting where needed.

Who uses it
Customer and supplier support
What is needed
Access to the support channel and to each customer’s history.

Entry service: AI implementation

The phone, at any hour Calls outside opening hours are lost and peak-hour calls get abandoned.

A phone agent that resolves, opens a ticket and books a slot when a person is needed. PA · Phone Agent · See it on the platform

  1. Step no longer done by hand: Outside hours it goes to voicemail The call is answered at any hour
  2. Step no longer done by hand: At peak hours the call is abandoned Whatever can be resolved gets resolved
  3. Next day nobody knows what they wanted The rest reaches you with a ticket open

Work per call, in this example

8 min 2 min

What a person still does

Handling what needs a person.

Who uses it
Whoever answers the phone
What is needed
The phone system and wherever tickets get opened.

Entry service: AI implementation

Content, channels and catalogue

Publishing stops depending on someone having time, and the brand stays consistent in every piece.

The weekly newsletter The send depends on one particular person having time.

The send generated with your own templates and lists, loaded into your email tool.

  1. Step no longer done by hand: Someone hunts for something to write about The content is generated with your templates
  2. Step no longer done by hand: They build the piece in the tool The send sits loaded and ready
  3. The send slips to next week You reread it and press send

Work per send, in this example

4 h 45 min

What a person still does

Approving the content and sending it.

Who uses it
Marketing
What is needed
The templates and access to the sending tool.

Entry service: AI implementation

Posting on social Every post is improvised and the brand loses consistency.

Project templates with the visual system settled, and the content generated inside them.

  1. Step no longer done by hand: Every post is improvised the same day Project templates settle the visual system
  2. Step no longer done by hand: Whatever template is at hand gets used The content is generated inside them
  3. The brand loses consistency post by post You pick what gets published

Work per post, in this example

40 min 10 min

What a person still does

Picking what gets published and when.

Who uses it
Marketing and whoever runs the channels
What is needed
The brand’s visual system, even if it is only a PDF.

Entry service: Training

Imagery and creative Every campaign needs a photography budget and three weeks.

Image generation and variation over your own brand and your own product. PI · Product Imagery · See it on the platform

  1. Step no longer done by hand: A photo shoot gets booked Images are generated over your own brand
  2. Step no longer done by hand: You wait for the shoot and the retouching Variants come out of a single piece
  3. Every new variant costs again You pick which ones get used

Work per campaign, in this example

20 h 3 h

What a person still does

Picking the image and approving it.

Who uses it
Marketing and catalogue
What is needed
Product photography and the brand’s visual rules.

Entry service: AI implementation

Product listings The catalogue grows faster than it can be described.

Attributes pulled from the photo, copy written and categorised against your own tree. GP · Generative Products · See it on the platform

  1. Step no longer done by hand: Someone looks at the photo and writes the copy Attributes are pulled from the photo
  2. Step no longer done by hand: They categorise it by hand against the tree The listing comes out written and categorised
  3. New products pile up undescribed You review before publishing

Work per listing, in this example

20 min 3 min

What a person still does

Reviewing and publishing.

Who uses it
Catalogue and ecommerce
What is needed
The photos and the category tree you already use.

Entry service: AI implementation

Several languages Translating the catalogue costs more than that market brings in.

Catalogue translation in bulk, published rather than left waiting to be reviewed one by one. GP · Generative Products · See it on the platform

  1. Step no longer done by hand: It gets sent out for translation in batches The whole catalogue is translated in one pass
  2. Step no longer done by hand: You wait weeks and pay by the word The brand glossary applies itself
  3. The review stays pending You review a sample

Work per language, in this example

40 h 4 h

What a person still does

Reviewing a sample and publishing.

Who uses it
Catalogue and international expansion
What is needed
The catalogue in an exportable format and the brand glossary.

Entry service: AI implementation

If one of these cases is yours, write to us

Technical teams

Daily work inside the technical team.

Building software

AI goes inside the cycle — requirements, prototype, change, review — not into a tool on the side.

Executable requirements The requirements document leaves out exactly what has to be asked later.

Requirements with paths, acceptance criteria and tests, written for an agent to execute.

  1. Step no longer done by hand: The document gets written in prose The requirement carries paths and acceptance criteria
  2. Step no longer done by hand: What is missing surfaces while implementing The tests are written from the start
  3. What is not written gets asked over chat You decide what is in scope

Work per requirement, in this example

2 h 45 min

What a person still does

Deciding the scope and validating the result.

Who uses it
Whoever writes the requirements and whoever implements them
What is needed
Access to the repository and an hour from whoever knows the domain.

Entry service: Training

The right tool for the job Everything is built in the same language because it is the one the team is fluent in.

Each piece in whatever suits it, with the agent covering the learning curve.

  1. Step no longer done by hand: The language the team is fluent in gets chosen Each piece goes in whatever suits it
  2. Step no longer done by hand: The piece gets forced where it does not fit The agent covers the learning curve
  3. The cost shows up in maintenance You decide the architecture

Work per new component, in this example

16 h 6 h

What a person still does

Deciding the architecture and holding it up.

Who uses it
Technical leads
What is needed
Openness to revisiting a technical decision already made.

Entry service: Training

From requirement to prototype The argument happens over a document and the mistake surfaces once it is built.

A clickable prototype before deciding, built in hours and discarded at no cost.

  1. Step no longer done by hand: The argument happens over a document The clickable prototype comes out in hours
  2. Step no longer done by hand: The real version gets built straight away It gets tried with whoever will use it
  3. The mistake surfaces once it is done You decide whether it gets built

Work per prototype, in this example

40 h 6 h

What a person still does

Deciding whether it is worth building.

Who uses it
Product and whoever decides
What is needed
Somewhere to publish it and someone to look at it.

Entry service: AI implementation

Assisted development The team repeats the same pattern file by file.

An agent inside the repository, with your own conventions as its context.

  1. Step no longer done by hand: You look for how something similar was solved The repository context is already in view
  2. You write the change and the tests You write the change with tests alongside
  3. Step no longer done by hand: You repeat the same pattern file by file The pattern is applied in one pass

Work per change, in this example

6 h 2 h

What a person still does

Deciding how it gets solved.

Who uses it
The development team
What is needed
Access to the repository and the team’s conventions written down.

Entry service: Training

Code review and quality Review depends on who happens to be available that day.

A first review that applies your own guidelines, not the model’s generic ones.

  1. Step no longer done by hand: The change waits for someone to look at it A first review applies your own guidelines
  2. Step no longer done by hand: Review depends on who is around that day The obvious things arrive already fixed
  3. The criteria live in each person’s head You review what actually needs deciding

Work per review, in this example

1 h 15 min

What a person still does

Deciding what the guideline does not cover.

Who uses it
Whoever reviews and whoever submits changes
What is needed
Integration with the change tracker and the review criteria written down.

Entry service: AI implementation

Keeping knowledge when people leave

What the team knows ends up written and in use, not only in the memory of a few people.

Living documentation Documentation goes out of date because nobody reads it and nobody updates it.

Documentation that is the agent’s configuration: if it is wrong the agent fails, so it stays right.

  1. Step no longer done by hand: Documentation gets written at the end The documentation is the agent’s configuration
  2. Step no longer done by hand: Nobody reads it and nobody updates it If it is wrong the agent fails and it shows
  3. It goes out of date and stops being reliable You fix it when it fails

Work per release, in this example

4 h 20 min

What a person still does

Writing the criteria that are not in the code.

Who uses it
The whole technical team
What is needed
Access to the repository and somewhere to publish it.

Entry service: AI implementation

The legacy system Nobody wants to change the old system and whoever built it has left.

A catalogue of what it does, a dependency map and a plan to take it apart in pieces.

  1. Step no longer done by hand: Nobody remembers what each part does A catalogue of what it does comes out
  2. Step no longer done by hand: Every change is made blind The dependencies end up mapped
  3. It ends up being left alone You decide what gets taken apart first

Work per module, in this example

24 h 6 h

What a person still does

Deciding what comes apart and in what order.

Who uses it
Whoever inherited the maintenance
What is needed
The code, even with no documentation and nobody who remembers it.

Entry service: AI implementation

Access, operations and governance

The business queries without going through you, and you know who queried what.

Infrastructure and deploys A deploy means jumping between the repository, the cloud and the cluster.

An agent with read access over GitLab, AWS and Kubernetes, and write access only on confirmation.

  1. Step no longer done by hand: You jump between repository, cloud and cluster State arrives read from all three places
  2. Step no longer done by hand: State gets checked by hand in three places The routine steps get prepared on their own
  3. The deploy depends on who knows how You confirm before anything is written

Work per deploy, in this example

45 min 10 min

What a person still does

Confirming every write.

Who uses it
Whoever deploys and whoever is on call
What is needed
Read-only credentials to begin with, inside your environment.

Entry service: AI implementation

Monitoring with a cause Failures are discovered when a customer reports them.

A scheduled check that separates healthy from broken and explains the cause, not just the symptom.

  1. Step no longer done by hand: The alert fires without saying why The check separates healthy from broken
  2. Step no longer done by hand: The logs get read by hand The cause arrives explained, not just the symptom
  3. Often the customer notices first You decide whether to act

Work per incident, in this example

3 h 1 h

What a person still does

Deciding whether to act and how.

Who uses it
Operations and on-call
What is needed
Access to the logs and metrics you already collect.

Entry service: AI implementation

Data for the business The business cannot query without asking IT, and handing over the database is not an option.

An MCP with per-role permissions and a defined access order, running inside your environment.

  1. Step no longer done by hand: The business opens a ticket per question The business asks and the MCP answers
  2. Step no longer done by hand: Someone in IT writes the query Each person sees only what is theirs
  3. The result gets exported and emailed Only what it cannot answer reaches you

Work per query, in this example

30 min 2 min

What a person still does

Defining who may see what.

Who uses it
IT, to shrink the request queue
What is needed
Deciding who may see what. We put in the rest.

Entry service: Custom development

Agents connected to your systems An agent with no access to the systems cannot answer with real data.

A connection to the ERP, the CRM and the document manager, API first and SQL as the last resort.

  1. Step no longer done by hand: The agent cannot see the company’s data The agent queries the ERP and the CRM
  2. Step no longer done by hand: The context has to be pasted in by hand The context arrives on its own, with its permissions
  3. It answers generically and gets corrected You check what it comes back with

Work per request, in this example

40 min 5 min

What a person still does

Checking and deciding on what it returns.

Who uses it
The teams already using those systems
What is needed
API or read access to the ERP, the CRM and the document manager.

Entry service: Custom development

Internal process automation Routine operations still need a technician and a console.

Execution from plain language, with confirmation required before anything is written.

  1. Step no longer done by hand: Every routine operation opens a ticket The operation is asked for in plain language
  2. Step no longer done by hand: A technician goes into the console The agent lays out what it is going to do
  3. The same step repeats every week You confirm before it writes

Work per operation, in this example

25 min 3 min

What a person still does

Confirming the run.

Who uses it
Internal support and operations
What is needed
Access to the ticket tracker and permissions on the systems involved.

Entry service: AI implementation

Governing the agents Nobody knows what the agents query, what they cost or what happens when they fail.

Access rules, traceability, context limits and secrets kept out of the repository.

  1. Step no longer done by hand: Nobody knows what the agents query Every query stays logged
  2. Step no longer done by hand: The cost shows up on the invoice Access rules apply themselves
  3. Secrets end up in the repository Secrets live outside the repository

Work per monthly review, in this example

4 h 30 min

What a person still does

Defining the rules and reviewing them.

Who uses it
Whoever answers for the platform
What is needed
Somewhere to deploy it and whatever data policy you already have.

Entry service: AI implementation

If one of these cases is yours, write to us

eCommerce and CRM

What decides whether a customer buys again.

Enriching the CRM with predictive models

The CRM stops only holding what already happened and starts saying what each customer will do.

Unified data Every system says something different about the same customer.

A single profile with value, recency and propensity, synced back into your systems. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: Every system says something different about the same customer The systems come together into a single profile
  2. Step no longer done by hand: It gets cross-checked by hand in a sheet Value and propensity are worked out on their own
  3. The cross-check is stale the moment it is done The profile goes back into your systems

Work per monthly cross-check, in this example

10 h 30 min

What a person still does

Deciding what to do with the profile.

Who uses it
IT and the business, together
What is needed
Access to the systems that hold each piece of the customer today.

Entry service: AI implementation

What each customer is worth, and will buy The whole base gets worked the same way and budget is split on instinct.

Lifetime value, consumption cycle and purchase propensity models written into the CRM itself. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: The base is segmented by what it already bought Every record carries its lifetime value
  2. Step no longer done by hand: Budget is split evenly across it Propensity by category is worked out on its own
  3. What a customer is worth is estimated on instinct You decide where the budget goes

The return, measured

+30%

in incremental sales, with the models wired into their CRM operation

See the Norauto project

What a person still does

Deciding the strategy for each segment.

Who uses it
Marketing and sales leadership
What is needed
Transaction history and write access to the CRM.

Entry service: AI implementation

Anticipating which customers will leave Customers leave without warning and it shows up at month end.

Churn and value prediction, with the list of who to call this week. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: The churn shows up at month end Churn risk is worked out every week
  2. Step no longer done by hand: You call whoever someone remembers The list arrives ordered by value
  3. Nobody knows who is worth keeping You call the ones worth calling

Work per monthly review, in this example

6 h 30 min

What a person still does

Calling, and deciding what to offer.

Who uses it
Sales leadership and retention
What is needed
Transaction history with date and customer.

Entry service: AI implementation

Querying the customer base Every question about the base becomes a data request and a week of waiting.

An agent that queries the unified profile and returns the customer list with its reasoning in view. AA · AI Analytics Assistant · See it on the platform

  1. Step no longer done by hand: The question goes to the data team The question is asked in plain language
  2. Step no longer done by hand: You wait for the report The list comes out with its reasoning in view
  3. The answer lands once the campaign has gone out You decide whether the campaign goes out

Work per query, in this example

5 h 10 min

What a person still does

Deciding what to do with the list.

Who uses it
Marketing and sales leadership
What is needed
The unified profile and read permissions by role.

Entry service: AI implementation

Automations by purchase cycle and preference

Every send goes out when it is that person’s turn, not when the marketing calendar says so.

What to offer, and when The same campaign to the whole base returns less every time.

Segments and recommendation over the unified customer profile, at the moment they open it. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: The campaign goes out the same to the whole base Every customer lands in their own segment
  2. Step no longer done by hand: What to offer is decided on instinct The recommendation comes from the unified profile
  3. The result gets looked at when the month closes You decide which campaigns go out

Work per campaign, in this example

8 h 2 h

What a person still does

Deciding the offer and the message.

Who uses it
Marketing and sales leadership
What is needed
Purchase and contact history, even if it is spread around.

Entry service: AI implementation

Writing when it is time to restock The send calendar is the same for everyone and matches nobody.

Automations fired by each customer’s own consumption cycle, not by the date on the calendar. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: The send goes out on the day the calendar says Every customer has their cycle worked out
  2. Step no longer done by hand: Everyone gets the same thing on the same day The send fires when it is that person’s turn
  3. Someone who just bought gets the same offer You sign off the message and the rules

The return, measured

+25%

in engagement across their CRM campaigns

See the Norauto project

What a person still does

Writing the message and setting the rules.

Who uses it
CRM and marketing
What is needed
Purchase history with dates, and the automation tool.

Entry service: AI implementation

What each subscriber sees The newsletter carries the same six products for the whole list.

One-to-one recommendation inside the send template, resolved the moment it is opened. DM · Customer Datamodels · See it on the platform

  1. Step no longer done by hand: Someone picks six products for the whole list Every subscriber gets their own selection
  2. Step no longer done by hand: The same block goes out to everyone The block resolves when the email is opened
  3. Clicks are measured against the total You decide the template and the rules

The return, measured

+44%

in revenue from personalised CRM communications

See the Condis project

What a person still does

Deciding the template and what gets pushed.

Who uses it
Marketing and CRM
What is needed
A catalogue with price and stock, and the sending tool.

Entry service: AI implementation

An agent that writes with the history in view Personal follow-up only reaches the customers somebody has time to look at.

An engagement agent that writes by email, chat or SMS with the customer profile and propensity in view. EA · Engagement Agent · See it on the platform

  1. Step no longer done by hand: Follow-up reaches whoever there is time for The agent prepares each customer’s conversation
  2. Step no longer done by hand: The message is written from scratch every time The message goes out with the profile in view
  3. It gets written without the history in view You answer what the agent escalates

Work per conversation, in this example

25 min 5 min

What a person still does

Answering what escalates, and setting the tone.

Who uses it
Marketing and support
What is needed
The unified profile and access to the sending channel.

Entry service: AI implementation

One-to-one personalisation in real time

The store arranges itself differently for every visitor, at the moment the page loads.

One-to-one recommendation in the store Every visitor sees the same home page and the same featured products.

A one-to-one recommendation neural network, resolved in real time on the home page, the product page and the cart. RT · Real Time · See it on the platform

  1. Step no longer done by hand: The home page shows everyone the same featured products Every visitor sees their own selection
  2. Step no longer done by hand: The blocks are built by hand every season The blocks resolve as the page loads
  3. A first-time visitor sees the best sellers You set the business rules

The return, measured

+12%

in average order value for people getting personalised recommendation

See the Condis project

What a person still does

Setting the business rules and what gets pushed.

Who uses it
eCommerce
What is needed
A catalogue with stock, and the store’s browsing events.

Entry service: AI implementation

The order of the listing and the search The listing comes out in the same order for everyone and what is relevant sits three screens down.

One-to-one ranking of listings, search and best sellers, worked out for each visitor. RT · Real Time · See it on the platform

  1. Step no longer done by hand: The listing comes out in the same order for everyone The order is worked out for each visitor
  2. Step no longer done by hand: What is relevant sits three screens down What is relevant moves to the first screen
  3. The best seller buries the rest of the catalogue You set what always gets priority

The return, measured

x4.68

better ordering quality than a best-seller strategy

See the Condis project

What a person still does

Setting what always gets priority, and what does not.

Who uses it
eCommerce
What is needed
A catalogue with stock, and the browsing events.

Entry service: AI implementation

An agent that helps people choose People who do not know what to look for browse without finding it and leave without buying.

A store agent that asks, narrows down and recommends over the real catalogue and the visitor profile. EA · Engagement Agent · See it on the platform

  1. Step no longer done by hand: Someone who does not know what to look for browses without finding it and leaves The agent narrows the search down in the same visit
  2. Step no longer done by hand: The question ends up in the contact form The recommendation comes from the real catalogue
  3. The answer arrives the next day You handle what the agent escalates

Work per query, in this example

12 min 2 min

What a person still does

Handling what escalates, and reviewing the reasoning.

Who uses it
eCommerce and support
What is needed
A catalogue with stock, and your sales reasoning written down.

Entry service: AI implementation

If you sell to returning customers, write to us

Before starting

What is needed to implement a case.

Access to what you already have

We work on your systems and with your permissions. Nothing to migrate and no tool to replace.

Your criteria, in writing

How that work is done well. A short description is enough: without it, AI does the generic thing.

Someone who knows the process

The person who knows why it is done that way. They do not need a technical background.

One case first

One, until it is working. The next ones are decided with that one in use, not in a meeting.

Is one of these cases yours?

Tell us which one and we will tell you what it would take to implement it. And if it is not worth it, we will tell you that too.

Subject: What we implement — one of these cases is ours

We use your details to answer your request. Data controller: PLEASE NETWORKS, S.L. You can exercise your rights and read the detail in our privacy policy (opens in a new window).

Write from the full form

No commitment You talk to the person who will do the work If it does not fit, we tell you