Skip to content
L&L Synergy
Get in touch

HomeBlogAI agent examples

AI agents · In practice AI agent examples by industry: six typical processes with trigger, tools and approval

There are plenty of general explanations of AI agents. What is missing is a look at everyday work in an industry: what triggers the agent, what does it access, who approves? Here are six typical processes from the trades, tax advice, real estate, online retail, wholesale and hospitality.

Updated October 2026 · 5 min read · L&L Synergy

In short

  • All examples follow the same pattern: an event triggers the agent, it reads and looks things up in your systems, prepares a result, and a person approves.
  • The examples are typical processes, not client projects. What brings most for you depends on where the most time is lost today.
  • According to Eurostat, around one in five EU enterprises with ten or more employees used at least one AI technology in 2025.
  • An agent should not take over binding commitments, advice with liability or decisions about people.

Six processes at a glance

Industry Trigger and task Access to Who approves
Trades New enquiry with photos: the agent asks for missing details, proposes a site visit and creates a draft quote with a materials list Mailbox, calendar, price list, quoting software The master craftsperson approves the quote
Tax advice Documents arrive by email or portal: the agent sorts them, assigns clients, spots missing documents and prepares posting proposals Mailbox, client folders, accounting software The specialist checks postings and deadlines
Real estate Enquiry about a property: the agent answers standard questions from the listing, asks questions about the prospective buyer or tenant and arranges viewings Listing, calendar, prospect list The estate agent leads the viewing and negotiates
Online retail Customer question “Where is my order?” or a return: the agent checks status in the shop and with the carrier, replies and issues a returns slip Shop system, shipping data, returns rules Refunds above a fixed amount and special cases go to people
Wholesale and industry Enquiry with an item number or drawing: the agent identifies the item, checks stock and lead time and prepares the quote ERP, item master, price lists Sales approves the quote
Hospitality Group enquiry: the agent checks occupancy, replies in the guest’s language and creates a draft offer Reservation system, menus, calendar The management confirms

More on individual processes in the articles Automate customer enquiries and Automate quotes with AI. A short AI consulting engagement settles which process comes first in your business.

One process in detail: an enquiry at a trades business

A typical process, not a client project:

  1. Intake: An enquiry arrives by email: “Heating is dripping, photo attached.”
  2. Understanding: The agent reads text and photo, recognises the type of system and the urgency, and finds the customer in the customer database. On keywords such as water damage it informs the on-call person at once.
  3. Follow-up: If something is missing, such as the age of the system, it creates a follow-up question as a draft.
  4. Appointment: It checks the calendar, proposes two site visits and creates the case in the job management software.
  5. Quote: After the visit it prepares a quote from your notes and the price list.
  6. Approval: The master craftsperson checks, amends and approves. Only then does the quote go out.

What all examples have in common

  1. A clear trigger: an enquiry, a document, a message. The agent does not wait for questions; it reacts to events.
  2. Reading instead of guessing: the agent takes data from your systems, not from the model’s general knowledge.
  3. A result to check: a draft, a proposal or a summary, with sources.
  4. A person at the end: nothing goes out and nothing is posted without someone having checked. Only after weeks of good results is approval relaxed for simple cases.

What an agent should not take over

  • Binding commitments: people decide on prices, deadlines and goodwill. The agent supplies the draft.
  • Advice with liability: tax, legal or medical assessments stay with specialists.
  • Decisions about people: selecting job applications and assessing creditworthiness count as high-risk under Annex III of the AI Act. More under risk categories.
  • Unlimited rights: OWASP names “Excessive Agency” and prompt injection as central risks. An email can contain hidden instructions. So the agent gets only the rights it needs, and write actions go through an approval.

How to find the right process

  1. Collect five candidate processes: where does someone jump daily between mailbox, spreadsheet and software?
  2. Rate them on three criteria: how often does the process occur? How much time does it cost? Is there a clear result a person can check quickly?
  3. Write the process down: trigger, steps, data, exceptions and who approves.
  4. Start in test mode: the agent runs alongside the existing work and nothing goes out without a check.
  5. Measure: time per case, error rate, share of cases a person had to correct.

Sources

  1. Eurostat: Use of artificial intelligence in enterprises (2025 data)
  2. Anthropic: Building effective agents
  3. OWASP: Top 10 for LLM Applications 2025
  4. Regulation (EU) 2024/1689 on artificial intelligence, Annex III: EUR-Lex

Frequently asked questions

Are there ready-made AI agents for my industry?

There is industry-specific software with AI features. Whether it fits your processes and systems depends on the case. A tailor-made agent pays off when your processes deviate from the standard or several programs have to be connected.

How long until an agent is up and running?

That depends on the process, the data and the interfaces. After an assessment of the process we can give you a plan. More important than speed is a test phase in which you check the results.

Do I need my own servers?

Not necessarily. What matters is a data processing agreement, servers in the EU and clear access rights. For particularly sensitive data, running it in-house can make sense.

Does an agent replace employees?

The examples show something else: the agent does the groundwork such as reading, gathering and drafting. Checking, deciding and relationship building stay with people.

Which process should run first at your business?

Tell us briefly what it is about and which systems are involved. We will get back to you personally with an assessment, free of charge and without obligation.

Get in touch

Read on

AI agents for business AI use cases for business: 12 examples Have AI agents developed AI reporting and report automation