Welcome to Lampi AI Documentation

Starting with a new AI platform is not always easy. This documentation is your go-to resource, covering everything from initial setup to advanced use of Lampi.

Here, you will find practical advice and step-by-step explanations for Lampi’s features, helping you unlock their full potential in your daily work.

We designed this documentation to be both informative and accessible, so you can find answers to your questions and discover new ways to work with Marc and his specialist agents.

Discover more

Explore helpful tips, best practices, and detailed guides for using Lampi’s capabilities in ways that match your team’s business needs.

Not sure where to start?

Lampi is designed for teams with different levels of AI experience. Use the path below to begin with the right level of guidance, then learn by applying Lampi to real work.

👤 For users

These guides will help you get set up and begin using Lampi in your daily work.

Start by meeting Marc and his agents. You will learn how Marc suggests useful tasks, coordinates specialist agents, runs approved Routines, and returns work for your review.

New to AI?

If you have limited experience with generative AI tools, begin with the basics. Lampi is designed to be approachable, but a few good habits will help you get clearer, more reliable results.

📚 Read the documentation

Begin with the core product guides and best practices:

• Meet Marc and his agents.
• Understand Chat Agents and Grid Agents.
• Learn how to work with the AI Chat Agent.
• Review common mistakes to avoid.
• Learn how to write effective requests.

How you describe a task affects the result. Give Marc a clear objective, the right context, the expected output, and any important constraints. You do not need technical vocabulary or a perfect prompt: start with the work as you would explain it to a colleague, then refine the request in conversation.

🧪 Experiment with basic features

Begin with a small, reviewable assignment. Ask Marc to summarize a document, prepare a meeting brief, compare a short list of companies, or answer a question from approved context.

Review the result and try adjusting your request. This will help you understand how the selected context, instructions, and examples influence the output.

🎯 Learn through hands-on practice

The best way to learn is by doing. Start with work you already understand well, so you can judge the quality of the result. Move gradually from one focused assignment to larger work with Grid Agents and repeatable work with Routines.

💬 Ask for help

Contact the Lampi support team when you need help choosing an agent, connecting a source, setting up a Routine, or improving a result.

For experienced AI users

If you already use generative AI tools, you can move quickly through the basics and focus on Lampi’s connected, repeatable, and collaborative workflows.

🔨 Build and test

Create focused Chat Agents, structure work at scale with Grid Agents, and turn recurring assignments into Routines. Test each workflow on representative examples and define what a review-ready result should contain.

🤝 Share and collaborate

Share effective agents, Routines, examples, and review practices with colleagues. Reusable team standards help everyone begin with stronger instructions and consistent expectations.

⚙️ Customize the way work gets done

Connect approved business context, configure specialist agents for your team, and use Living Work Memory where useful so later assignments can build on relevant decisions and history.

Build progressively

Start with a task you can review easily. Once the result is reliable, apply the same approach to more inputs with a Grid Agent or make it repeatable with a Routine.

🔐 For administrators

Begin by setting up the organization, access, and shared context your team needs.

Task	What to do
Invite members	Send invitations from the administration space and confirm that each person joins the right organization.
Create roles and permissions	Give each member access appropriate to their responsibilities and the sensitivity of the work.
Manage approved context	Organize shared sources so relevant teams can use them while restricted information remains protected.
Set review boundaries	Define where human review or approval is required before sensitive or external use.

This foundation makes it easier for teams to use agents and Routines consistently without losing control of access or final decisions.

🚀 Build a team of internal champions

Successful adoption requires people who understand both the team’s work and the opportunities created by AI. Identify champions in each department where Lampi is deployed.

Strong champions usually:

• understand the team’s recurring workflows and deliverables;
• are comfortable experimenting or want to develop their AI skills;
• look for practical ways to improve quality and efficiency;
• help colleagues review results and adopt effective practices;
• share clear feedback with administrators and the Lampi team.

Responsibilities of a champion

• Create agents and Routines: configure Chat Agents, Grid Agents, and repeatable workflows for the team’s work.
• Document use cases and templates: maintain examples, instructions, and review criteria that have worked well.
• Train colleagues: help team members understand how to request, review, and improve AI-assisted work.
• Share success stories: show where Lampi improved speed, consistency, coverage, or the quality of a deliverable.
• Gather feedback: collect recurring questions, limitations, and improvement ideas from the team.

💡 Support a successful rollout

• Create a dedicated hub: maintain a shared channel or workspace with approved agents, Routine templates, examples, and guidance.
• Organize regular sessions: review new use cases, compare results, and agree on improvements together.
• Share and train: provide practical training based on real team workflows and review standards.
• Measure useful outcomes: track time saved, additional coverage, output quality, and adoption rather than the number of prompts sent.
• Keep people in control: make review responsibilities clear, especially for sensitive, external, legal, or investment-related work.

note

This guide will continue to evolve as Lampi adds new capabilities and teams develop better ways of working with Marc and his agents.

Next: Meet Marc and his agents →