tecnovy AI Suite

Developing AI Agents

Your agent takes on a task from a business process, breaks it into steps and calls tools you defined yourself in Python. It pulls data from existing systems through APIs or the Model Context Protocol (MCP) and grounds its answers in company knowledge through retrieval. Before critical steps it pauses until a human approves the next action. You build this agent part by part during the workshop and check it against test cases you write yourself.

Attendance certificate4.8/5 on ProvenExpert

What this module delivers.

From demo agent to one working through a business process step by step: tecnovy's Developing AI Agents suits people who read and adapt Python scripts. Afterwards you build your team's agent prototypes yourself, with tests and cost in view.

How the training runs

Method: One continuous Python project that grows part by part
Basis: tecnovy curriculum: agent patterns with and without a framework
Outcome: tecnovy certificate of participation, no examination

Dates & booking

Choose a date that fits

No public dates are scheduled yet.

New dates are published regularly. Join the waitlist or plan an in-house course for your team.

Request in-house training

Fit

Who this module is designed for

Typical roles

You build software or automations and are asked to implement processes with AI agents.
You work in data or IT and build prototypes that connect language models to internal systems.
You have built agents with no-code tools and now need your own tools, branching or tests.

Prerequisites

No prerequisites

You can start right away. What you need: basic Python skills, basic experience with APIs, basic knowledge of AI agents such as from the AI Agents Foundations course.

Consider instead AI Agents Foundations For people who do not program: you learn how agents work and which processes suit them, and put first agents together without code.

Curriculum

Developing AI Agents in Detail

Five parts, one continuous Python project: you start with the agent loop and add tools, knowledge and memory, multi-step flows and tests part by part. Systems of several agents, protection against attacks and production operation are left to AI Agents Deep Dive.

01Architecture of an agent

You set up the skeleton your agent grows on during the workshop.

The agent loop: plan, act, observe
Agent frameworks at a glance and when you can do without one
The same pattern with a framework and directly through a language model's API

02Tools

You give your agent tools it uses to act in existing systems.

Defining your own tools: name, description, parameters and return value
Connecting existing systems through their APIs
Connecting systems through the Model Context Protocol (MCP)

03Knowledge and memory

You decide what your agent knows and what it remembers.

Bringing in company knowledge through retrieval
Short- and long-term memory
Managing context: what the model sees at each step

04Multi-step flows

You turn single calls into a flow that also handles failures.

Planning and branching
Error handling when a tool fails
Implementing human approvals in code

05Testing and measuring

You check what your agent does and what it costs to do it.

Writing test cases and assessing results
Logging calls and tracing failures
Measuring cost and run time per run

Outcome

What you will be able to do afterwards

01
You explain the agent loop of planning, acting and observing, and decide whether a framework or your own code fits better.
02
You define your own tools with a name, description, parameters and return value that a model calls precisely.
03
You connect your agent to existing systems through APIs and MCP.
04
You bring in company knowledge through retrieval and control what stays in the agent's short- and long-term memory.
05
You build multi-step flows with planning, branching and error handling.
06
You implement human approvals in code wherever the agent should not decide alone.
07
You write test cases for your agent and assess its results systematically.
08
You log your agent's calls and measure the cost and run time of each run.

tecnovy Teilnahmezertifikat

Confirms your attendance and the curriculum covered in the training. Issued by tecnovy, shareable on LinkedIn.

Open the Certificate Showroom tecnovy →

Offer scope

What is included

Included

Project repository with a reference solutionThe reference solution shows the workshop's finished agent and serves as a template for your own agents afterwards.

Why tecnovy

What you get on top with us

No Slideshow, Hands-On!

Promised: no PowerPoint marathon. We work in groups, tie theory to practice, and you get real project examples from our experienced trainers plus the exchange with like-minded people.

Certificate Showroom

Get your certificate of participation and, if you have one, add your exam certificate from E-Learning. Fully automated, beautifully designed. Just for you, only at tecnovy.

Flexible Date Change

If you are not able to attend the course, you can rebook your training free of charge up to one week before the start of the training.

Attend Twice, Pay Once

You are welcome to attend the training online again within a year as a refresher.

Learn from Experts

We always guarantee you the use of didactically and methodically first-class qualified trainers who draw their knowledge from training experience as well as professional practical and project experience.

FAQs

Frequently asked questions

01Which framework does the workshop use?

tecnovy's Developing AI Agents works with an established agent framework and shows the same patterns without one, directly through a language model's API. That way you see what a framework does for you and can decide whether your own project needs one.

02How good does my Python need to be?

For tecnovy's Developing AI Agents it is enough if you can read and adapt simple Python scripts. You also need basic experience with APIs, because your agent calls existing systems through them.

03What does tool use mean for language models?

Tool use, also called function calling, means you describe a tool with a name, a description and a JSON schema for its parameters; the model replies with a structured call, and your code runs it and returns the result. That is how Anthropic and OpenAI describe it in their API documentation (as of November 2026). In tecnovy's Developing AI Agents you define such tools yourself.

04What do the exercises need technically?

The exercises in tecnovy's Developing AI Agents call a language model through its API and need API access with a key for that. Anthropic, for example, bills such calls by the tokens processed, tool descriptions included (as of November 2026), which is why the workshop also has you measure what one run of your agent costs.

05How does the workshop differ from Developing Your Own MCP Server?

tecnovy's Developing AI Agents works on the agent side: your agent connects existing systems through APIs and MCP. Building an MCP server for an internal system yourself is the subject of the tecnovy workshop Developing Your Own MCP Server.

06How does the workshop differ from AI Agents Foundations?

tecnovy's AI Agents Foundations needs no programming: there you assess use cases and build first agents with no-code or low-code tools. In Developing AI Agents you write agents in Python, with your own tools, retrieval, memory and test cases.

07Is the agent ready for production after the workshop?

No, tecnovy's Developing AI Agents produces a prototype. You learn what to check on it: test cases, error handling, human approvals, logs and cost per run.

08Is there an examination or a certification at the end?

No, Developing AI Agents is a tecnovy workshop with no examination and no certification. If you attend at least 80% of it, you receive the tecnovy certificate of participation.

09What comes after Developing AI Agents?

The next step at tecnovy is AI Agents Deep Dive: systems of several agents working together, protecting them against attacks such as prompt injection, and running them in production with monitoring. These topics are deliberately left out of Developing AI Agents.

What does your training at tecnovy look like?

AI Agents DevelopmentDeveloping AI Agents

No dates yet

We will let you know as soon as the next date is scheduled.