AI assistants: finding the right expertise in Basel
In Basel's pharmaceutical sector, life sciences and Rhine logistics, some requests require expertise spread across professions and international R&D sites. The managerial question is specific: can a no-code assistant route a request to the right contact without multiplying handoffs?
What the literature establishes
Samer Faraj and Lee Sproull (2000, Management Science) show, in software development teams, that expertise coordination is associated with performance beyond the mere presence of expertise within the team. This coordination includes knowing where expertise resides, recognising when it is needed and bringing it to bear. For assistant design, this suggests distinguishing two functions: producing an answer and finding someone who can handle the request.
A directory is not a collective memory
Kyle Lewis (2003, Journal of Applied Psychology) develops and validates a measure of transactive memory systems organised around specialisation, credibility and coordination. The mechanism therefore goes beyond knowing everyone's areas of expertise: it also involves confidence in that expertise and the way people work together. Applied to an assistant, this distinction cautions against treating a match between keywords and job titles as effective routing.
The premature conclusion
One might conclude that connecting an assistant to the staff directory is enough to improve access to expertise. These studies establish neither the effectiveness of such a tool nor its ability to recognise a person's availability, remit or situated knowledge. A plausible referral remains a proposal: the recipient must be able to confirm that the request actually falls within their remit. (our executive and employee training programmes)
What the evidence does not establish
These studies concern human coordination and team memory, not no-code generative assistants. They provide design criteria, but no estimate of the benefit an assistant would bring to a Basel organisation. Counting only the referrals generated would be particularly misleading: the tool could speed up the first handoff while increasing subsequent transfers.
A practical check in Basel
Within SHR's programme « Créer ses assistants et agents IA sans coder », a relevant exercise would be to prototype a routing assistant for non-sensitive requests between pharmaceuticals, life sciences, Rhine logistics and international R&D. Before and during a pilot of equal duration, record, for comparable types of requests, the proportion for which the first recipient confirms having both the expertise and the remit to handle them, together with the number of reroutings needed to obtain that confirmation. Keeping an identifier, the successive recipients and their confirmation for each request makes the measure verifiable without storing technical content. The proposed criterion is better first-recipient routing without an increase in reroutings, rather than a rise in the number of answers produced by the assistant. To go further: explore the Building AI Assistants and Agents Without Coding training in Basel in the canton of Basel-Stadt, or browse our executive and employee training programmes in Switzerland.
In pictures: Building AI Assistants and Agents Without Coding in Basel
- building ai assistants and agents without coding
- Basel
- research
- training basel