Custom AI for Swiss SMEs, enterprises and the public sector

AI agents to reduce paperwork and speed up your organization

Tenders, invoices, reports and case files, live in weeks and hosted in Switzerland. We are a Zürich team of ETH and EPFL engineers: we start with one workflow, prove it in a fixed-price pilot and then run it in production with you.

  • Built in Zürich
  • Your data stays in Switzerland
  • You own the code

Invoice agent

Running

  1. Read 42 supplier invoices from the inbox
  2. Matched each one to its purchase order
  3. Posted 39 invoices to the ERP
  4. 3 amounts differ from the order: waiting for your review
Illustrative example of an agent run.Finance
01 / 25

Start small

Three steps, each with a result, a duration and a price

You do not buy a program. You buy one step at a time, and you decide after each one whether to take the next.

  1. 1 day

    Step 1: AI workshop

    A ranked list of use cases with a build-or-buy recommendation for each.

    from CHF 4’500

  2. 4+ weeks

    Step 2: Pilot

    A working agent on your real data, measured against the manual baseline.

    CHF 2’000 to 10’000

  3. Custom

    Step 3: Production and operation

    Integration into your systems, hosting in Switzerland and monitoring.

    Fixed quote and monthly fee

Not sure which step is yours? Start with a free 30-minute discovery call.Book a call

See the steps in detail

Use cases

Seven kinds of paperwork an AI agent can take over

Each one starts the same way: one workflow, a fixed-price pilot, and a comparison with how the work is done today.

  • Supplier invoices

    An agent reads invoices from your inbox, matches them to orders and prepares the posting in your accounting system. A person reviews the cases it flags.

  • Tenders and offer comparison

    An agent pulls the requirements out of the tender documents and compares the incoming offers item by item, so your team decides instead of copying numbers into tables.

  • Inspection and technical reports

    Notes, measurements and photos from the field become a finished report in your template, ready for the engineer to check and sign.

  • Questions about standards and regulations

    Staff or members ask in plain language and get an answer that shows the exact passage it comes from.

  • Customer advice and recommendations

    An assistant that knows your product range and your rules helps advisers and customers find the right product.

  • Replacing Excel and manual reporting

    One tool with one source of data replaces the spreadsheets that get emailed around, and the monthly report builds itself.

  • Visual checks with a camera

    A camera and a trained model check parts, shelves or installations and report only the cases a person needs to look at.

Where our team built its craft

Prior experience of our founding team. Trained at ETH Zürich and EPFL.

Proof, not promises

Outcomes we have delivered

in deal value won via cloud architectures we designed for RfPs
CHF 100k
users positively evaluated our 3 pharma AI PoCs, all selected for pilot rollout
30+
content-creation time, cut from days, with applied GenAI
<1h
of manual in-store inventory checks removed, per store each week
~4h

Our promise

A working pilot from four weeks

This is what we commit to when your pilot starts. It is a promise for your project, not a statistic from past ones.

Our work

Real systems, in production, measured against outcomes

We build for the organizations that keep Switzerland running: finance, healthcare, public administration, and beyond.

Industries

Who we build for

The paperwork differs by industry, the way we start does not: one workflow, a fixed price, hosting in Switzerland.

  • Manufacturing and industry

    Manufacturing and Industrial Firms

    Supplier invoices, offer comparison, inspection reports, and visual checks for mid-size industrial firms.

  • Associations and standards bodies

    Associations and Standards Bodies

    Questions about standards and regulations, member inquiries, consultations, and member reporting.

  • Multi-site service groups

    Multi-Site Service Groups

    Invoices, group reporting, inspection reports, and visual checks for groups that run many locations.

  • Public sector

    Public Sector

    Small fixed-price mandates: a plain-language writing assistant, a reporting tool that replaces Excel, case files.

Research

What the engineers at Anthropic and OpenAI learned

The newest finding from the frontier labs’ engineering writing, and what it means for your systems. We follow this work closely so your build starts from what holds up today.

Trusted: deterministic code
Untrusted: sandboxed
GatewayConfigures the agent, separate container
Credential brokerThe agent never sees secrets
Auto reviewA second agent approves each action
LogEvery config change and every turn
AgentIts own identity
Workspace and networkOwn directories, allowlisted hosts
configuresusesevery action

Hundreds of agents, none trusted

OpenAI has “persistent agents deployed across the entire company, including hundreds of OpenClaw agents”, and “over 500 agents that run proactively in the background doing meaningful work”. The agent and the code it writes are treated as untrusted and sandboxed; trusted, deterministic code owns networking, identity and secrets. A second agent reviews every action (“we enforce auto review on all agent actions”), and “we log everything”.

What it means for you

Agents that run on their own inside a company are safe because of the walls around them, not because the model behaves: a sandbox, a credential broker so the agent never sees secrets, its own identity, a reviewer for every action and a full log. We build those walls first.

Source: OpenAI, “Securing persistent agents like OpenClaw at OpenAI”, 12 November 2026. Our drawing of the post’s text.

Patterns we build with

How modern AI systems are actually built

This field changes every few months. These are the patterns, protocols and mental models we bring into every client engagement, drawn from what the frontier labs publish and from what holds up in production.

Agent patterns

The agent loop

An agent decides its own next step. It gathers context, acts, checks the real result and repeats, with stopping conditions and human checkpoints to keep control.

Source: Anthropic, “Building effective agents”

  1. Goalfrom a person
  2. Gather context
  3. Take actioncall a tool
  4. Check resultground truth
  5. Done

Repeat until the task is done, a limit is reached or a person is needed

1 / 7

Legend

  • Model or agent
  • Tool or system
  • Data
  • Check
  • Person
See all patterns

Automation and augmentation

Automating today’s work is only the tip of the iceberg

The first instinct of most organizations is to hand their existing workflows to AI. That is worth doing, but it leaves most of the value on the table. The real gain is the work you have never been able to do, because it needed people or budget you did not have.

Automation

Do today’s work faster and cheaper

  • Takes a workflow you already run and does it end to end
  • Saves time and cuts errors
  • The value is capped by the work you already do
  • Where almost every organization starts

Augmentation

Do the work that was never feasible

  • Tasks that were too expensive, or needed people you never had
  • Intelligence on tap, whenever and as much as you need it
  • Adds new value instead of only trimming cost
  • Where frontier AI models create the most value

Work that becomes possible

  • Read every contract you have signed, not a sample
  • Check every invoice, not one in fifty
  • Answer every customer or citizen request the same day
  • Follow up on every lead and every open offer
  • Prepare a briefing before every meeting
  • Track every new tender, regulation and competitor move

We help you find the work that is now within reach, and build the agents that do it.

Find your hidden workflows

Why Super AI Labs

Not another subscription. Not another slide deck.

Our founders spent years delivering inside large consultancies and Swiss banks. We started Super AI Labs to do the part that actually helps: build the system, hand it over, and stay accountable for it.

Off-the-shelf AI tools

Made for everyone, fitted to no one

  • One generic product for every company
  • Another per-seat monthly subscription
  • Your process bends to fit the tool
  • Cost grows with headcount, value often does not
  • Your data lives inside someone else’s product

Super AI Labs

Small, senior, hands-on

  • Built around your process, inside your tools
  • Yours to keep: code, system and IP. No per-seat pricing
  • A working pilot from four weeks, at a fixed price
  • The engineers you meet are the ones who build
  • Measured on outcomes, with human review where it matters

Large consultancies

Big, expensive, slow

  • Months of analysis before anything runs
  • Slide decks instead of working systems
  • Day rates that pay for overhead
  • The people who sell are not the people who build
  • Method libraries instead of measurable results

Swiss by default

Compliance and data residency, built in from day one

Your data stays in Switzerland, and we design around the rules our clients answer to. Not as an afterthought.

  • Your data stays in Switzerland

    Swiss hosting with data residency by design, plus logging and backups chosen to match.

  • Compliance-aware from day one

    We design around nFADP (revDSG), GDPR, and sector rules like FINMA and GxP from the start.

  • Local team, your language

    Engineering, hosting, and support from Zürich, in German or English.

Our own products

We also ship software of our own

A regular agency builds software for other companies. We do that, and we also build, launch and run our own products. It keeps our engineering sharp and proves we can take software from zero to real users.

  • GastroBro

    Launching 2026

    Financial performance software for restaurants, bars, and nightclubs.

  • FarmClub

    Launching 2026

    A Booking.com and Airbnb hybrid for innovative US farm stays.

  • TradingView Magical Keyboard Shortcuts

    Live · ~300 weekly users

    Custom keyboard shortcuts for professional TradingView users.

Ready when you are

Which workflow should we start with?

Show us where your team loses time. In 30 minutes you will know whether AI can take it over, and what it would take to get there.