PD Training

Enhancing Course Customization and Efficiency with AI Technology

Introduction

PD Training, a leader in professional development solutions, has implemented an innovative AI-enhanced course recommendation system to streamline course customisation, improve client engagement, and integrate seamlessly with existing technological infrastructure.

This is not your average “rainbows and unicorns” case study that many companies produce. Instead, this will explore what happens when ambitious but very achievable goals mix with misaligned expectations, communication, and outcomes.

Project Overview

PD Training sought to leverage artificial intelligence to enhance its course offering processes, aiming to deliver personalised course recommendations that precisely meet client needs. The project was designed to integrate with PD Training’s existing Azure environment and custom-built CRM, ensuring a smooth workflow and maintaining high data privacy standards.

Key Features of the AI Model

  1. Integration with Current Systems: The AI model was developed to align seamlessly with PD Training’s existing Azure environment and CRM, facilitating a unified workflow and data consistency.
  2. Custom Course Suggestions: Utilizing AI Search technology, the model analyses client inputs to suggest tailored course outlines, effectively combining modules from various disciplines such as time management, leadership, and communication skills.
  3. Data Utilization: The AI was trained on PD Training’s extensive library of course outlines suitable for AI processing, enhancing the model’s learning capability and recommendation accuracy.
  4. Scope Control and Content Privacy: The model was designed to provide recommendations strictly within the scope of PD Training’s offerings, using internal data sources only to ensure content privacy and integrity.
  5. Enhanced User Interaction: Capable of handling initial client inquiries, the LLM (AI large language model) interacts with the user to help guide and encourage deeper decision-making processes and a focus on the customer’s strategic outcome.
  6. Testing and Feedback: Regular testing and feedback to continuously refine the AI model, ensuring its effectiveness and adaptability to changing client requirements.

Project Implementation

The project was rolled out over a seven-week period, following a structured timeline that included phases for development, integration, testing, and deployment. Each phase was marked by rigorous evaluation and stakeholder feedback to ensure the system met all operational requirements and performance benchmarks.

Project Issues

For CROFTI, it seemed like a fairly straightforward AI bot implementation project. It has several specialised technical steps and knowledge that we’ll implement, but nothing that appears too difficult. The source documents that the AI would draw its “truth” and content from needed a little massaging, but the internal resources at PD Training were well positioned to make the changes quickly and as needed – something that’s normally our biggest roadblock was not the case here.

Instead, the primary issue that revealed itself halfway through the project was the misaligned expected outcome versus the scope of works signed off during the sales process. The key person who would be “signing off” on the results of this bot was not involved in the initial sales conversations, nor the project kick-off, or detailing sessions at the start of the project.

It’s Project 101 – stakeholder identification and engagement! For us, we didn’t pick up on who the right stakeholders were to be in the meetings at the start. The result meant that after the initial 4-5 weeks of bot building, testing, and prototyping, we ended up with something quite different to what was going to be accepted.

Eventually, we managed to meet with the right person to guide the project, which required a significant rework of our approach. The impact was felt across all aspects:

  • How the bot would initially interact with a customer was wrong. It needed to be hyper-focused on the discovery of the customer’s underlying reason for engagement
  • How the bot would present the results and information, between the chat interface, and separate “improvement blueprint” window
  • The performance of the bot – speed and presentation of responses
  • Capture of user details – where and when best to capture this information
  • Context and language the bot uses

While none of these were technically complicated, the deliverables had shifted from where they began. As professionals brought in to ensure a successful outcome, we did not push for proper clarity from the right people. A lesson learned the hard way!

Impact, Lessons, and Benefits

Impact

The project’s impact was a mix of learning experiences and unmet expectations. While the AI-enhanced course recommendation system reached a functional state according to the initial requirements, it was ultimately not deployed for client use. PD Training, committed to maintaining its reputation for professionalism and polished service, decided that the bot required further refinement before it could be presented to customers.

This decision underscored the importance of aligning project outcomes with client expectations, particularly in maintaining brand integrity. Although the bot met the technical specifications, it fell short of the operational polish needed for customer-facing interactions.

Lessons

  1. Stakeholder Identification and Engagement: The critical lesson here was the importance of involving all key stakeholders from the outset. The absence of the primary decision-maker led to a significant disconnect between what was delivered and what was expected. Ensuring that all relevant parties are engaged early and throughout the project is essential.
  2. Clear Scope Definition and Communication: This project highlighted the need for precise and transparent communication regarding the scope and expected outcomes. Misalignments early in the project led to rework and delays, which could have been mitigated with clearer initial agreements and ongoing dialogue.
  3. Technical and Functional Understanding: Despite the project’s challenges, the team gained valuable insights into the technology behind AI Search and LLM-based chatbots. This increased understanding will be beneficial in future projects and contribute to more refined implementations.
  4. Flexibility and Adaptability: The ability to adapt to changing requirements and feedback, though challenging, was crucial. This flexibility allowed us to pivot and reassess our approach to better align with the client’s needs, even if the final product wasn’t immediately deployable.

 

Benefits

  1. Enhanced Technical Expertise: The project significantly boosted their team’s knowledge and expertise in integrating AI Search and LLM technologies. This learning curve will enhance the future capability to deliver more sophisticated solutions they may progress with.
  2. Process Improvement: The experience prompted both parties to refine project management practices, particularly in stakeholder engagement and scope definition. These improvements will help mitigate similar issues in future projects.
  3. Professional Development Mindset: The project reinforced the importance of accurate communication and realistic expectations in experimental technology implementations. Both PD Training and CROFTI have grown from this experience, fostering a culture of continuous professional improvement.

Conclusion

This case study highlights the complex reality of implementing cutting-edge technology. While the project did not culminate in a customer-ready solution, it provided invaluable lessons and opportunities for growth. The experience emphasized the importance of clear communication, stakeholder engagement, and adaptability. At CROFTI, we remain committed to supporting PD Training as they continue to explore and refine their technological capabilities, ensuring that future initiatives are better aligned with their high standards of professionalism.