Easier, faster, more efficient: The integration of an AI chat into a laboratory information system (LIMS) simplifies interaction, speeds up the retrieval of searched content and noticeably increases overall efficiency.
The AI solution and its area of application
Traditional laboratory information systems (LIMS) are usually highly complex and require specific technical knowledge and extensive training in order to be used effectively. Acceptance by employees and the training of new recruits are also success factors that should not be underestimated. The answer to these typical challenges in everyday laboratory work can be the integration of an AI chat, such as ChatGPT. This enables users of a LIMS to access complex data structures easily and directly, perform analyses for samples via natural language interaction and interact with the LIMS anywhere in the laboratory using mobile devices. This innovation is therefore of particular interest to laboratories facing an increasing demand for efficiency and digitalization, as well as those for which the shortage of skilled workers is likely to be a key challenge.
Efficiency gains and other benefits
The integration of an AI chat functionality improves the accessibility, user-friendliness and acceptance of a LIMS, as this creates an intuitive communication bridge between the user and the system. It also minimizes sources of error, optimizes training time for employees and increases overall laboratory productivity. In addition, there is currently an increasing use of AI chats such as ChatGPT in a wide range of everyday work applications. The tangible benefits in terms of time savings in research, knowledge transfer and decision-making are also evident when working with a laboratory information system – a benefit for the effectiveness of the laboratory and its attractiveness as an employer.
The technology in detail
The integration of an AI chat requires a LIMS with a correspondingly modern software architecture. An open web interface that allows data to be queried as well as created and changed is essential. One such modern and highly flexible LIMS is SAMPLES, developed by qualitype GmbH. Among other things, it has an innovative GraphQL interface that enables secure data exchange.
For the solution described here, the Large Language Model ChatGPT was integrated directly into the SAMPLES laboratory information system via an intuitive chat interface. The chat application has two interfaces: One to the LLM and one to the LIMS. The LLM can be used in the same way as the familiar ChatGPT. However, the application also recognizes special requests from the user during the conversation that relate to the LIMS. Corresponding functions are stored in the application for this purpose, which are then executed. Depending on how these are defined, users receive the desired information from the LIMS via the interface, for example by retrieving samples or results with certain filter criteria (“in which samples was batch XY used?”). However, actions can also be carried out directly via the application, such as creating new samples in the LIMS (see illustration). Consequently, the application can be used as an equivalent additional interface – with a multitude of additional advantages and application options.