Knowledge Orchestrator: AI Transforms Corporate Knowledge

Dr. Errol Brandt, following the unexpected loss of a key team member, recognized the critical need to safeguard corporate knowledge. The challenge was mitigating catastrophic knowledge loss, which often surfaces when crucial business processes are disrupted by incomplete understanding—especially when dealing with complex spreadsheets. This lack of accessible, deep technical knowledge complicated pricing and capital investment decisions, ultimately impacting business performance.

Driven by this need, Dr. Brandt initiated a passion project on IBM Cloud to develop a solution. By late 2022, his prototype was ready for testing. He then partnered with two co-founders to establish Knowledge Orchestrator Pty Ltd, with the goal of commercializing their innovative solution. Today, Knowledge Orchestrator uses its own platform to meticulously document every facet of its operations, empowering its team to maintain productivity from any location.

The core of Knowledge Orchestrator's platform is its unique PEG (pre-enhanced generation) approach. Unlike retrieval-augmented generation (RAG), which validates content from large language models (LLMs), PEG focuses on building high volumes of automated ground truth through a proprietary content generation algorithm developed on IBM Cloud using the IBM watsonx.ai AI studio. This platform transforms unstructured data into structured natural language, creating accessible ground truth. By segmenting knowledge into specific domains, the algorithms operate more efficiently, reducing inferencing costs and minimizing the risk of hallucination—major barriers to using generative AI in management analytics.

Knowledge Orchestrator's adoption of its own solution has yielded significant results. With a team of five FTEs plus KIRA, an AI assistant, the company's knowledge base has grown to over 2,000 articles (500,000 words). Each employee reads an average of 9.3 articles and contributes 0.9 articles daily, demonstrating high engagement driven by the platform's comprehensive documentation of business aspects. IBM's solution, facilitated by Meier Business Systems (MBS), addresses scalability, availability, and resource management by deploying inside a Kubernetes cluster on IBM Cloud. It also leverages the IBM Carbon Design System for a seamless user experience and runs its unique content generation algorithms within IBM watsonx.ai, utilizing IBM's cloud services and APIs for natural language processing and foundation models.

Value Results:

• 3.8x increase in employee productivity: AI-powered applications have significantly augmented employee output.

• 18.5 knowledge updates per employee per month: Demonstrates active engagement and contribution to the knowledge base.

• Enhanced Scalability and Availability: The solution's deployment within a Kubernetes cluster on IBM Cloud ensures future scalability and high availability.

• Improved Decision-Making: Accessible, structured knowledge enables better pricing and capital investment decisions.