Different results for the same process
Criteria, tone and format change depending on who runs it.
onext Enterprise AI
Hand a process to an AI assistant
Each assistant works with your company's documents. Where there is risk, a person approves.
If you want to develop with AI, go to AI-Accelerated Development.
- a month per assistant
- From €250
- per assistant and company
- Monthly fee
- the first two assistants
- Licia and Marco
The problem
Each team uses AI its own way
Sales, operations, marketing and finance already work with AI, and three problems keep coming back.
The AI doesn't know your company
For security reasons, sensitive documents should not be uploaded to a public tool. The answers come out generic: they know nothing about your catalogue, your internal rules or your exceptions.
What one person learns never reaches the rest
A good instruction or a good template stays with whoever created it. Everyone starts over.
The result: individual licences and nothing the company can reuse.
What onext Enterprise AI is
What you get with each assistant
Your documentation as context
An assistant can consult collections of your company's documents: rules, previous tender documents, guides.
A person approves where there is risk
In processes with risk, the assistant proposes and a person decides from a queue of pending tasks. An audit record is kept.
It works when you decide
You schedule a process to run on its own, for example a daily search, and it leaves the result for you to review.
Data kept separate per company
Each company has its own database, encrypted and isolated from other clients', and the AI models run in the EU.
How we work
How we work: one loop to deliver, one to improve.
It is onext's method, and the one we follow with you.
-
Delivery loop
Understand the problem.
We investigate before we build and define the idea with its edge cases. A person approves it.
-
Specify before we program.
What will be built is written down with criteria that can be checked. The specification and the design are approved before any code is written.
-
Build with checks and audit.
Agents build under automated checks, other specialised agents audit the work, and a person reviews and signs off on the change.
-
Improvement loop
Evaluate the AI with data.
When the product has AI inside, we measure it against reference cases and compare it with the previous version, with cost alongside.
-
Operate and improve.
We watch cost, quality and deviations. What we learn goes back to the start.
One loop to deliver, one to improve. A person decides where there is risk.
The assistants
Two assistants to start with
Each handles one specific process.
Public tenders assistant
Licia
Finds the public tenders that fit your company. It helps you prepare the response with your documentation, and a person reviews it.
Marketing assistant
Marco
Prepares digital marketing drafts, process by process. A person reviews them before they go out.
If you need another assistant, tell us in the conversation.
Price
You pay per assistant, not per user
You start with the process that hurts most and add others when it makes sense.
- Licia, €250 a month. Marco, €500 a month.
- An initial set-up. Its amount is agreed in the conversation, before any commitment.
To know more
Three reads on context, assistants and when to start.
Context engineering: why it will survive the agent Marketplace
A Q2 2026 pillar piece on the one layer the Marketplace can't sell. It separates the three layers of the enterprise AI stack (cognitive, control, context), explains why only the context layer resists commoditization, and breaks the discipline down into five operational layers: curated instructions, working memory, episodic memory, retrieval policy, and invalidation/expiration. It closes with the cost-per-request contract, the three most recurrent anti-patterns (RAG as a band-aid, prompt-as-database, memory with no TTL), and a brownfield migration plan executable in four weeks without stopping production. Read
Start with AI now or wait 6 months? The right question is a different one
The AI decision is the most deferred one in Spanish leadership committees in 2026. This piece separates the three legitimate reasons to wait (high regulatory risk with no legal analysis, data with no operational conditions, nobody with the judgment to lead) from the three reasons to start now (organizational learning curve, competitive window, talent). It brings the empirical record from Bain & Company (41% in production report positive results vs 25% in pilot) and the public Q1 2026 cases from European and American integrators. It closes with four contexts on each side, the two traps that invalidate the decision and the single question that frees the committee from the cycle of deferrals. Includes a downloadable one-page matrix to take to your next committee. Read
Sovereign AI is no longer data residency: the split of control in 8 axes
Q2 2026 pillar piece that closes the sovereign AI cluster. It separates the three concepts that get mixed in the Spanish and European market (infrastructure sovereignty, vendor-stack governance, operational sovereignty) and develops the split of control across eight operational axes: identity, data, model, context, agent, logging, human-review gates and exit. It includes the comparison matrix of onext against three market archetypes (industrial sovereign-AI brand, EU sovereignty integrator, Microsoft Agent 365 GA May 1 with the E7 bundle at $99/user/month). It closes with when you need each proposal and the exercise to take to your next RFP. ReadSee where your team stands
Answer 16 questions about how your team works with AI today. It is free.
Start the self-assessment- questions
- 16
- blocks
- 4
Contact
Tell us which process you want to hand over
We look at which process makes sense to hand to an assistant and whether Enterprise AI fits you.
What happens next
- You write to us. Two lines are enough.
- We reply and set up a call.
- On the call we tell you whether we can help and where to start.