Prompt Lab

Learn How to Instruct Gen AI Models

Master Prompt Engineering

Whether you’re a newbie or looking to supercharge your skills with advanced models, our Prompt Lab is your gateway to mastering prompt engineering. 

If you’re curious, ambitious, and ready to explore the cutting edge of Prompt Engineering, jump in!

Fill out the form and make sure that you stay ahead in these crazy days of super intelligent AI models.

Why Join Us?

Hands on

Say goodbye to boring lectures. We’re all about engaging, real-world applications that you can start using immediately.

Free

Zero cost, zero barriers. All you need is an internet connection and your enthusiasm.

Flexible

Can’t make a session? No problem! Catch up anytime with recordings available on YouTube.

INteractive

Get your questions answered in real-time during our live Q&A segments.

You’ve just found an Easter egg!

Upcoming Sessions

Boost your skills and learn how to craft effective prompts to communicate with AI models like a pro.

Join a community so that you are connected with like-minded individuals. Share your ideas and maybe even get feedback for your next big project.

Stay ahead and keep up with the latest trends in AI and make yourself stand out in your field.

Full Calendar

#1 - Mar 26th - Fundamentals: How to Write a Solid Prompt

Date and Time

Wednesday, 26.03.2025, 5pm – 6pm CET

Join us online:
meet.google.com/uih-okez-sip

Google Calendar

Content

Key Components of a Well-Structured Prompt:

  • Directive: Crafting clear instructions
  • Examples: Utilizing one-shot and few-shot prompting
  • Role (Persona): Assigning roles to guide AI perspective
  • Additional Information: Providing context and background
  • Output Formatting: Specifying desired response formats


Best Practices:

  • Clarity and conciseness
  • Use of action verbs
  • Positioning the directive for maximum effectiveness

Date and Time

Wednesday, 02.04.2025, 5pm – 6pm CET

Join us online:
meet.google.com/jwd-cyyy-vnj

Google Calendar

Content

Analyzing prompt effectiveness:

  • Ensure clarity and specificity of the prompt .
  • Check if the prompt provides sufficient context for the desired output.
  • Evaluate if the prompt clearly defines the task or objective.
  • Consider if the prompt uses appropriate prompting techniques (e.g., zero-shot, few-shot) for the task.


Analyzing response quality:

  • Evaluate the accuracy and factual correctness of the information provided . Catch potential hallucinations.
  • Assess the coherence and logical flow of the response .
  • Verify if the response adheres to the specified output format and length constraints.

Date and Time

Wednesday, 16.04.2025, 5pm – 6pm CET

Join us online:
meet.google.com/fzb-octt-cpr

Google Calendar

Content

Chain of Thought (CoT) Prompting

  • Simulating human-like reasoning by explicitly prompting the model to generate a step-by-step explanation.
  • Breaking down complex tasks into smaller, more manageable step.
  • Variations: Contrastive CoT, Faithful CoT, Automatic CoT 


Tree of Thoughts

  • Allowing the LLM to explore and evaluate multiple different reasoning paths in a structured manner, like a tree
  • Suitable for general problem-solving and scenarios requiring complex decision-making.


Self-Consistency

  • Generating multiple diverse reasoning paths for a single prompt and then selecting the most consistent answer among them.
  • Applying majority voting or consensus to identify the most frequent answer.


ReAct Prompting (Reasoning and Acting)

  • Combining reasoning with task-specific actions
  • Interacting with environments during problem-solving

Date and Time

Wednesday, 23.04.2025, 5pm – 6pm CET

Join us online:
meet.google.com/wsy-migx-tqw

Google Calendar

Content

Text: Use clear and specific instructions, provide context, specify format, and use examples .  

Images: Be specific in instructions, add examples, break down complex tasks, specify output format, and for single images, place the image first. If the model isn’t drawing information from the relevant part, provide hints. If the output is too generic, ask the model to describe the image first . 

Audio and Video: Utilize models that support these formats. Use plugins if necessary. Be specific with instructions about the content and desired analysis.

Date and Time

Wednesday, 30.04.2025, 5pm – 6pm CET

Join us online:
meet.google.com/ekd-ycyx-rhq

Google Calendar

Content

System Prompts

  • Setting overall AI behavior and capabilities
  • Defining roles and constraints


Deep Research Applications

  • Crafting prompts for in-depth research
  • Structuring prompts for complex queries


Prompting with Attached Files

  • Approaches for incorporating external content
  • File upload methods and content extraction

Don't Miss Any Session

Yah, we know, everything is available on YouTube. But it will only be live once. During the live sessions you can have your questions addressed and interact with others.

Need More Convincing?

Don’t miss out on this chance to supercharge your skills and be part of a vibrant community pushing the boundaries of what’s possible with AI.

Want to learn more about things we do?

Team member Aiku Data Engineering CEO Soraya Rosadha

Soraya

Host

Team member Aiku Data Engineering CEO Siegfried Eckstedt

Siggi

Facilitator

This website uses cookies to ensure you get the best experience.

ARE YOU IN?

Many subscribers already enjoy our premium stuff. Subscribe now.