Practical AI Workshop for Business Leaders

Leaders do not need an overview of dozens of AI tools. They need a safe way to apply AI to their own decisions, documents, and management processes. That is why half the program covers the fundamentals, while the other half focuses on participants' tasks.

What you get

Key Takeaways

Participants work confidently with Claude and other suitable models
The team distinguishes a one-off prompt from a repeatable agent workflow
The company receives a portfolio of tasks with clear value and complexity
Rules for working with confidential data are agreed before the practical work begins

Project journey

How the workshop works: from contract to working use cases

Before the first training session, we assess the team's level and actual processes, select priority operations, and prepare exercises using the client's materials.

Formalities

01
Contract and project kickoff
We agree on the format, price, and group composition, then sign the documents.
ResultScope, deadlines, and responsibilities are defined

Assessment

02
Participant level
We send a short questionnaire asking which AI tools participants already use, for what tasks, and where they encounter difficulties.
ResultWe assess the group's starting level
03
Workflows
We identify repetitive operations that take time and could potentially be delegated to an AI agent.
ResultWe receive a list of the company's actual tasks
04
Priorities
Together with the client's designated representative, we select the top operations and define the workflow.
ResultThe program is tailored to this team

Preparation

05
KT.Team preparation
We analyze operations, prepare instructions and quality criteria, and build demo versions of agents using test or anonymized data.
ResultPractice begins with prepared use cases
06
Meeting schedule
We agree on the dates, duration, and sequence of training sessions.
ResultParticipants know the plan and preparation requirements

Training and results

07
Practical workshop
Participants learn the general approach, then build and test agents for selected operations together with experts.
ResultThe team independently performs selected operations with the help of AI agents

Program

Sample workshop program

Module 1

Model capabilities and limitations

What modern models can do, where they make mistakes, which tasks can be delegated, and where the decision remains with a human.

Module 2

Task definition

Context, source materials, success criteria, workflow, and ways to verify the response.

Module 3

Leader's work

Documents, research, meetings, analytics, communications, and project control through practical examples.

Module 4

From request to agent

Sources, instructions, tools, access rights, human oversight, and saving the result in the working system.

Module 5

Company scenarios

The number of processes and depth of prototyping depend on the selected program format.

Module 6

Safe use

Data masking and anonymization, handling confidential information, logging, and agent publication rules.

Practical use cases

Seven scenarios to learn from

The case mix is tailored to participants' roles. For hands-on work, we choose tasks where a leader can quickly verify the quality of the result.

01

Meeting outcomes agent

Turns a transcript into decisions, tasks, owners, deadlines, and an email to participants.

02

Project monitoring agent

Collects status updates, compares plan versus actuals, and highlights deviations requiring a leader's decision.

03

Finance agent

Finds unpaid invoices, checks contract terms, and prepares reminders without manual reconciliation.

04

Document workflow agent

Checks the contract, calculation, and payment details, then prepares invoice and completion certificate drafts.

05

Basic accounting agent

Recognizes the document type, applies accounting rules, and routes ambiguous cases to a specialist.

06

Bank statement agent

Routes operations across legal entities, accounts, and 1C databases without mixing data.

07

Market research agent

Collects data, removes duplicates, creates a comparable report, and preserves links to the sources.

KT.Team experience

Workshops for a major developer's technical team

Delivered a program on applying AI in technical support and agent development. The team explored enterprise agent architecture, knowledge bases, tool integration, quality control, and escalation of complex cases to a specialist.

The company name is not disclosed under the confidentiality terms.

Formats

Three program levels

Choose the level of engagement: from a two-day workshop to launching your own agent lab.

Basic workshop

$3,600

The team learns to use AI and designs solutions for its own tasks.

2 three-hour sessions

Up to 15 participants
Preparation questionnaire
Specifications for 3–4 use cases
Recording and workbook
Personalized certificates

Choose basic

Recommended

Applied intensive

$5,300

Participants build their first prototypes and a portfolio of scenarios for future implementation.

3 three-hour sessions

Up to 15 participants
Assessment of up to 10 processes
5–6 use-case briefs
1–2 verifiable prototypes
Post-program consultation

Discuss the intensive

AI Agent Lab

$7,100

Training and a limited pilot environment for creating and testing agents.

4 three-hour sessions

Up to 15 participants
Technical preparation sprint
Pilot open-source environment
One published agent and 2–3 prototypes
Configuration, instructions, and administrator training

Request the lab

How we get started: 30 minutes to clarify objectives → participant questionnaire → program and fixed estimate → start upon approval.

Advanced format · $7,100

Pilot laboratory on an open-source platform

In the advanced format, we prepare a portable pilot configuration within 2–4 weeks. The team creates agents, connects approved knowledge, and tests use cases in a shared environment.

Agent builder

Instructions, workflows, and tools are brought together in one interface.

Corporate knowledge

Agents can be connected to an approved document database with access rules configured.

Model selection

We confirm model connectivity after a technical review of the selected platform and infrastructure.

Control

Roles, run history, errors, and confirmation of significant actions remain visible.

We choose the platform after reviewing its license, infrastructure, information security requirements, and available connectors. One candidate is openJiuwen Agent Studio, with visual agent building and process orchestration.

openJiuwen project on GitHub →

A controlled path for a corporate agent

Participant

Selects a processgoal and success criterion

Platform

Builds an agentinstructions, knowledge, tools

Control

Checks the resultsources, errors, verification

Work

Publishes the use caseaccess for selected roles

Only verified scenarios are published. Significant actions remain under employee control.

Pilot boundary. The price includes a preparatory technical sprint, a portable pilot configuration, deployment instructions, and one week of consulting support after handover. Deployment can take place in the customer's approved infrastructure; servers, models, integrations with 1C, CRM, ERP, and internal APIs, production-grade resilience, and operations are priced separately.

Experts

We assemble the team for the program

For the executive workshop, we involve experts in business processes and AI applications. The platform program includes an analyst and a technical lead.

Practical AI Workshop for Business Leaders

Andrey Putin

CEO of KT.Team

AI-native system architecture, technology change management, and connecting AI to business outcomes.

Lada Ikonnikova

Client Portfolio Manager

Process analysis, hands-on facilitation, and development of the company's AI initiative portfolio.

Dmitry Gusev

Lead AI Analyst

Designing agent use cases and requirements for knowledge, rules, and quality criteria.

Sergey Spitsyn

Technical lead

Agent solution architecture, integrations, access management, and the transition from prototype to a production environment.

Expert talks on YouTube →

Terms

Organizational terms

Tools: Claude can serve as the foundation. We also demonstrate other models and services when they are better suited to the task.
Account managers: Participants use their own subscriptions, or KT.Team helps arrange access to and payment for the services.
Recording: Online meetings are recorded and remain with the client.
Confidentiality: Before practice begins, we agree on data rules. We use masking, anonymization, and training copies of documents.
Certificates: After completing the program, participants receive personalized electronic certificates.
Group size: The optimal group size is up to 15 people. For larger groups, we create additional cohorts.