Exploitation of knowledge capital has been a frustrated goal in many large businesses. With the aid of AI, Starmind provides a low overhead, high impact means of leveraging enterprise knowledge resources.
Starmind is a powerful and pragmatic solution to the age old problem of how to find and share expertise within a large organization. Knowledge management and sharing has been a goal in many businesses, but the overheads of maintaining a knowledge base, and resistance from people who are already very busy,usually resulted in failure to share knowledge in a meaningful way. We can also add to this the fact that many people with valuable skills are not too keen to share all their hard earned experience with a knowledge database.
The user of Starmind simply enters a question in natural language through the interface. The internal algorithms identify the person most likely to be able to answer the question and sends an anonymous message. The expert then replies by sending a message back, possibly detailing other sources of expertise in the business. If the question has been answered before, then a response can be sent immediately. Obviously this is only going to work if issues such as workload balancing are addressed – and they are.
Within Starmind is an AI engine, and as with all AI the system needs some input, some algorithms, and to generate output. The input comes from users who need to know where to find expertise. The algorithms create a network – called a ‘Neural Know-how Network’, and this creates a representation of how topics are related to each other within a company. Based on Hebbian Learning rules, topics that often occur together grow together and build clusters. On the other hand, the more two topics occur separately, the less they are connected.
As well as the obvious knowledge of who-knows-what, other dimensions are added to the network. This includes time-to-resolution, availability of experts (local times for example) and overall resolution rate. Load balancing also means that a given expert does not get overloaded with questions.
Getting the knowledge network up and going is aided by algorithms that process existing content.
Some Use Cases
Customer facing service can obviously benefit from a platform such as Starmind. It means call center operatives get answers to queries faster – and hopefully happier customers. Project based work is also an obvious area of application, as is the process of collecting information for strategic decisions. As a Swiss company it should come as no surprise to find that Starmind has some large Swiss customers – Swisscom and Planzer being two good examples. Others include Ogilvy and UBS.