Developing the SLII® Chatbot™: Bridging Learning and Real-World Application
In leadership development, the gap between a positive classroom experience and practical application can be significant. Many learners leave training energized, only to feel overwhelmed by work and fall back into old habits, losing the momentum of their newly acquired skills. Recognizing this, we decided to incorporate an AI-enabled chatbot into Blanchard’s flagship SLII® leadership development program.
Our goal was clear: help learners sustain and apply their SLII® knowledge back at work by providing ongoing support, reminders, and tools. Thus, the SLII® Chatbot™ was born, developed in partnership with Mobile Coach.
Two Modes of Interaction: Structured and Unstructured
Chatbots generally fall into two categories:
Structured interactions: These follow prewritten scripts and predefined pathways, similar to an “enter 1 for tools” phone menu.
Unstructured interactions: Powered by large language models (LLMs), these allow natural, free-flowing conversations, such as what you’d experience with ChatGPT.
We wanted the SLII® Chatbot to offer both structured and unstructured capabilities, combining predictability and adaptability.
Building the Structured Framework
The structured part of the chatbot serves as a guide for learners during the critical first one to three months after training. We meticulously scripted dialog and prompts to help learners:
Stay engaged with SLII® concepts
Access tools and resources
Receive tips in a conversational and approachable tone
Creating the structured dialog was a familiar process for us, but required adapting to a text-based medium. Writing for chatbots meant embracing casual language, emojis, and abbreviated phrasing to align with modern communication styles. By the end, we had a robust 50-page script, blending practical support with a touch of fun.
Entering New Territory: The Unstructured Component
The unstructured element, powered by an LLM, was designed to answer any SLII®-related questions learners might ask, from “How do I approach an employee at D2?” to “What’s the best way to give feedback?”
To ensure data privacy and security, we collaborated with Mobile Coach to use a private LLM. This allowed us to feed proprietary SLII® content into the model while keeping it protected. The chatbot’s “training” included materials like facilitation notes, program documents, and supplementary resources.
The Art of Prompt Engineering
Designing prompts for an LLM is a crucial step in creating an effective chatbot. Prompts serve as the “instruction manual” for the AI, shaping how it interprets questions and generates responses.
Think of a prompt as a sandwich:
The user’s question is the meat.
Context and writing instructions are the bread and toppings that make the response complete.
For instance, if a learner asks, “I just had a disagreement with a team member—what should I do?”, adding context might look like:
“The user recently completed SLII® training and is trying to use the model to resolve a workplace conflict.”
You can also include writing instructions to shape the tone and style, such as:
“Write a concise, supportive response (under 100 words) that validates the user’s concern and references the SLII® concept of Diagnosing Development Levels.”
By layering context and clear instructions, we improved the quality and relevance of the chatbot’s responses.
Lessons Learned and Best Practices
Here are a few takeaways from our experience:
Content is King
The chatbot’s effectiveness hinges on the quality and quantity of the source material. Providing rich, well-organized content ensures the bot has a strong foundation to draw from.Clarity and Consistency in Prompts
Writing clear and detailed prompts is essential for shaping responses. Define tone, style, length, and context to achieve the desired output.Iterate and Improve
AI systems thrive on refinement. By analyzing user interactions and identifying gaps in the chatbot’s responses, you can continually enhance its performance.
The Future of AI in Leadership Development
The SLII® Chatbot represents a bold step into the future of AI-enabled learning support. While still early days, the potential for LLM chatbots to extend the value of training programs is immense. If you’re considering developing a similar tool, know that success requires patience, iteration, and a commitment to the long game—but the results are worth it.
We’re excited to see how this technology helps SLII® learners stay engaged and empowered, ensuring their leadership skills make a lasting impact in the workplace.
Best Regards,
Yogesh
For more details, visit our website: https://byldgroup.com/
Or call at: 1800-102-1345