Company

Success together

ATESO stands for Advanced Technology Solutions: We design and develop digital solutions with sustainable added value — driven by a passion for technology.

Who we are

Digital solutions with sustainable added value

As a Swiss software company based in Frauenfeld, we develop digital solutions with sustainable added value.

Usability
User-friendly from the ground up: We align the requirements of providers and users with the right technology.
Collaboration
Trust and friendly collaboration shape our work — with customers, employees and partners.
Speed
Our experts bring new concepts, software solutions and digital products to life in a short time.

Projects delivered: +20
Industries: +5
Years of expertise in the team: +12
Email response time: 24 h

Our AI Journey

Our AI Journey — There from the Start

The way software is developed and consulting is delivered is undergoing a fundamental shift — at a pace our industry has rarely seen before.

We have actively shaped this development from the very beginning, consistently aligning our way of working with an AI-native approach. We experiment systematically, challenge existing methods, and integrate the insights we gain into our products and services. What proves effective in practice becomes the standard; what fails to deliver sustainable value is decisively discarded.

This allows us to continuously evolve our approach, staying at the forefront of technology and methodology — without our clients having to bear the risks of this transformation.

For Context

Many companies use AI today, but only a minority achieves measurable economic impact. The bottleneck is no longer the technology itself, but how roles, processes, and decisions are rethought and reorganized.

We have navigated this transformation ourselves — and bring this experience directly into our work with clients.

How We Work in the AI Era

Tools — we call them by their name

Generative AI and specialized multi-agent systems in daily development, open-source models in our own operation, custom skills, agent rules, and fine-tunings where standards fall short.

We develop AI-native and call our tools by their name.

Generative AI, multi-agent systems, AI toolkits, agents, and skills in development
Open models in self-hosted operation: vLLM, Ollama, LM Studio, and Ray.io — locally or on our own infrastructure
Custom models, fine-tunings, and algorithms where standard models are not enough
In consulting and daily work: Microsoft Copilot, Gemini, and Google AI Studio

We don't just use models, we operate them.

Organization — the development cycle is agentic

From developer to architect. From specialist to reviewer. AI changes not only how we work, but also how roles, responsibilities, and collaboration are shaped within the organization.

01
Agentic cycle instead of code completion
AI is not just an autocomplete tool for us; it is used throughout the entire development cycle. Architectural decisions and approvals remain with us.
- Analyze the codebase and propose implementation approaches
- Coordinate changes across multiple files
- Check security, performance, and team conventions in reviews
- Keep documentation and changelogs up to date
02
The developer becomes an architect
Our developers plan across modules, services, and system boundaries instead of generating code line by line — they set the framework in which the machine works.
- Rules define what an agent may do and where approval is mandatory
- Patterns make solutions recognizable and thus verifiable
- Architecture determines whether today's speed costs tomorrow's maintainability
AI reinforces discipline, it does not replace it
Because agents produce faster, CI/CD, test automation, and code review have become stricter for us. Without these foundations, you only create chaos faster.

Speed and quality can only go hand in hand if the human sets the framework and the machine works within it — not the other way around.

Compliance & Responsible Use

Binding guidelines with training, model selection based on data classification, tailored permissions per agent — and in-house model operation for sensitive data.

Binding Guidelines with Training
We have clear rules on which content may run on which models — conveyed through training, not a distributed PDF.
Model Selection Follows Data Classification
The models used depend on the data protection level in accordance with nDSG and GDPR — whether an external provider or in-house operation is used follows from this.
03
Permissions and Tools Tailored to Each Use Case
What an agent does not need to do, it must not be able to do. Each use case receives exactly the permissions required for its task — no shared user accounts, no administrator accounts.
- Minimal toolset: read before write, no delete or send access
- Data access enforced in the source system, not via prompt instructions
- Existing permissions are inherited, not bypassed
- External actions require approval, every tool call is logged
In-House Model Operation for the Highest Protection Level
What must not leave the building is processed on our own infrastructure — the same capability that ExeoAI is built on.

If you are not allowed to see information without AI, you cannot see it with AI either — and what an agent does not need to do, it must not be able to do.

ATESO MWK-Stack

Our learnings are embedded in the ATESO MWK-Stack: a modular development environment for backend, web, and mobile — used in customer projects and in our Exeoflow.

More about the ATESO MWK-Stack

ExeoAI

An AI framework based on open models with custom fine-tunings — hosted on Swiss servers or on-premise at the customer, running continuously for sensitive data.

More about ExeoAI

For our customers, this means: the speed and cost-efficiency of the AI era — combined with the quality and reliability of a Swiss partner. Those who do not develop AI-natively today cannot compete in terms of speed and cost. We can do so without compromising our Swiss-made quality standards: economically viable thanks to open source and AI-native development.

We implement data protection technically and organizationally; the legal assessment of individual cases remains with your legal or data protection advisor.

Our Values

Six principles that define our work

Six principles define our work — from the first conversation to ongoing operations.

Passion

We inspire — with our passion for what we do.

We question the status quo and pursue innovations with passion. We identify our customers' needs and seek solutions that make them more successful.

Quality

Your digital solution — our commitment.

Quality in software development is non-negotiable for us. We create not just solutions, but value — from conception through deployment and beyond.

Efficiency & Effectiveness

We think long-term and work on consistent solutions.

We collaborate across departments with colleagues and partners, focus on our core competencies, and act purposefully and process-oriented — with clear communication of the next step.

Responsibility

We think in solutions, not problems — and we stick to our agreements.

We take the initiative to address matters in the best interest of the company. By delegating responsibility, we also grant our team members the corresponding competencies and our trust.

Reliability

We lead by example.

We are a competent, reliable partner for customers, suppliers, and colleagues. We communicate deviations promptly and transparently, and we complete orders on time and within the agreed scope.

Communication

We communicate openly and honestly — but concisely.

We share our experiences and solutions in constructive exchange with customers and colleagues. We address difficult situations directly to find solutions together.

What we do

From idea to operation

From technology consulting to cloud services, and individual software, web, and mobile development: We accompany you from the idea to operation — AI-native, cost-effective, and in Swiss quality.