Skip to content
noeta

SOLUTIONS

What can AI do in your own work?

Each solution starts from a task your teams already do. Read them as examples, not as a closed list: if your own case is not here, describe it to us.

Where does your data go? You choose

Most solutions can run in any of the three ways, and you can combine them: sensitive work locally, general work on a frontier model. Where the choice is narrower, the solution says so.

  • Cloud model

    A frontier model, for example Claude, through its API, with your data handled under a data processing agreement.

  • Anonymised

    Personal data is replaced with placeholders before the request leaves your network, and put back in the answer.

  • Local

    Open models on your own servers or a private cloud in the EU. Your data never reaches an outside AI service.

Everyday work

Documents, email, meetings, translation and reports: the work every team does each week, which makes it a good place to start.

  • Documents from your templates

    Proposals, offers, letters and first drafts of contracts, written from your approved templates, price lists and past work. A person checks and sends.

    For sales, legal and administration.

  • Inbox and request triage

    Sorts incoming email and requests by topic and urgency, drafts replies for approval and passes each one to the right person.

    For customer service, front desks and shared mailboxes.

  • Meetings to minutes and tasks

    Turns a recording or notes into minutes, decisions and tasks with owners, in Greek or English.

    For management and project teams.

  • Greek and English in your own words

    Translation and writing in your terminology and tone, from a glossary your team keeps.

    For marketing, export and customer communication. Runs in the cloud or locally.

  • Reports from your data

    Monthly and weekly reports from spreadsheets and exports of your ERP or CRM, with charts and a short commentary a person reviews.

    For owners, finance and operations.

Documents and data

What someone reads and keys in by hand today: invoices, tender documents, contracts, years of archives. AI reads them, and a person checks what matters.

  • Data from invoices and forms

    Reads invoices, delivery notes, applications and forms, and fills your ERP or spreadsheets, with checks and a person approving what looks wrong.

    For accounting, logistics and back office.

  • Tender analysis

    Reads the documents of a public or private tender, lists requirements, deadlines and required documents, and drafts the compliance table.

    For construction, IT and suppliers to the public sector.

  • Contract review against your terms

    Compares an incoming contract with your standard terms and marks every deviation for a lawyer to decide.

    For legal departments, procurement and law firms. Runs anonymised or locally.

  • Knowledge assistant

    Answers staff questions from policies, manuals and contracts, and shows the passage behind every answer.

    For HR, operations and technical teams.

  • Archive search

    Search across years of documents by meaning, not only by words, with every result linked to its page.

    For law firms, engineering and public bodies.

Agents and connections

For teams that want AI to work inside the systems they already use. A person approves each change.

  • Claude connected to your systems

    Your team asks Claude, and Claude reads and drafts in your CRM, ERP, website or ticketing system, with each person's own permissions and every change saved as a draft for a person to approve. This is how our own site works.

    For any team that already uses Claude. Runs in the cloud. Built on MCP, the open standard Claude uses to connect to tools.

  • Claude projects and skills per role

    Ready instructions, templates and reference files for each role, so the sales team, the accountants and the lawyers each start from their own setup.

    For companies on Claude for teams. Runs in the cloud.

  • Workflow agents

    Multi-step work done by agents, with a person approving each step that matters. Example: a new customer file, from collecting documents to checking they are complete and preparing the record.

    For operations, onboarding and compliance.

  • Assistant for your website visitors

    Answers visitors from your own content, in Greek and English, and passes the conversation to a person when it should.

    For service businesses, e-shops and public-facing organisations. Runs in the cloud or anonymised.

Privacy and compliance

A law firm or a defence organisation may need to anonymise everything before it reaches a frontier model, or to run its own agents locally. We build both, along with the records GDPR asks for.

  • Anonymisation gateway

    Finds personal and sensitive data in Greek and English text (names, tax numbers, social security numbers, addresses, case numbers, health details), replaces it with placeholders before the request reaches a frontier model, and puts it back in the answer. Every request is logged.

    For law firms, healthcare, HR, insurance and banks.

  • Local AI on your own servers

    Open models running on your servers or a private cloud in the EU. Your data never reaches an outside AI service.

    For defence and security, law firms, the public sector and research.

  • A team of local agents

    Several agents working together on your infrastructure, for example one that reads case files, one that checks them against rules and one that drafts the summary, with no outside service involved.

    For defence, legal and regulated industries.

  • Routing by sensitivity

    Sensitive work goes to the local model, general work to a frontier model, by rules you set and can audit.

    For companies that want both quality and control.

  • AI use policy and GDPR records

    A written policy on which data may go to which AI tool, the data map and records for your AI uses, and input for the data protection impact assessment, prepared with your DPO or legal advisor. Access control and audit logs for the tools we build. We prepare the material. The legal opinion stays with your DPO or lawyer.

    For any company using AI with personal data.

Seminars and training

Seminars for owners and workshops for teams, with one aim: your people doing their day-to-day work with Claude more easily and faster.

  • Seminar for owners and management

    What AI can do in your company today, what it costs, where the risks are and where to start, with examples from your sector.

  • Hands-on Claude workshops for teams

    Your people work on their own tasks with Claude: writing, summarising, analysing spreadsheets, research, preparing documents. Everyone leaves with something they will use the next morning.

  • Tracks by role

    Separate sessions for sales, finance and accounting, legal, HR and marketing, built around the tasks of each.

  • Safe use of AI

    What data may go into AI tools, how to check an answer, and your company's own rules. It also covers the AI literacy that the EU AI Act asks companies using AI to ensure for their staff (Article 4).

  • Follow-up

    Each team leaves with its own Claude projects and templates, and a follow-up session a month later on what worked.

Examples by sector

  • Law firms: case files organised, research with names removed, and drafts from your own templates. mydiki, which we are building for lawyers, shows what this looks like in practice.
  • Defence and security: everything on your own servers, with no outside service.
  • Accounting offices: invoices and documents into your software, and client questions answered from your files.
  • Education: lairn and Daskalakos AI, two products we built ourselves, show what can be done for a whole sector.
  • Companies that bid in tenders: requirements and deadlines from the tender documents.
  • Any team that already uses Claude: training, a ready setup for each role and connections to your systems.

Questions before we start

  • We do not see our own case. What now?

    Describe it to us. The solutions here are examples: each one is built to measure, around the customer's own task.

  • Do we have to start with several solutions at once?

    No. The pilot is one solution, on your own data, at a fixed price. You decide on the next ones once you have seen the result.

  • Will we have to change our systems?

    We start from what you already use: your ERP, CRM, spreadsheets and email. The solutions read from them and fill them in. They do not replace them.

  • What if it does not work as we expected?

    That is what the pilot is for: you check the result on your own data before we move to rollout. After that we stay, to adjust the solution as your work changes.

  • Should we start with training or with a solution?

    Either is a good start. A seminar shows what is worth doing. A solution starts from a task you already have in mind, and training your people is part of the rollout.

THE FIRST STEP

Which of these fits your work? Write to us, or describe a task you do not see here.