Syensqo.AI, a division started only six months ago by the Belgian science and technology powerhouse Syensqo, has developed SyGPT, an advanced AI chatbot powered by Azure OpenAI Service. This project aims to enhance data access and operational efficiency within Syensqo, using different data sources within the company to answer questions via an internal chatbot. Syensqo.AI leverages Azure OpenAI Service, providing fast and efficient answers in a secure manner.
A rich 160-year history driven by innovation and science
The key capability is really accessing data in a secured manner. We’re going live with a chatbot capability that allows our colleagues to access data [...] securely.
Vincent Colegrave, Head of Syensqo.AI, Syensqo.AI
Internal SyGPT Chatbot created in just one quarter
The journey to create SyGPT was driven by the need for rapid deployment and immediate value extraction. According to Colegrave, "The key capability is really accessing data in a secured manner. We’re going live with a chatbot capability that allows our colleagues to access data from different applications, databases, and existing data storage services in the company securely."
This rapid development was made possible through Syensqo's strategic partnership with Microsoft using Azure OpenAI Service. "We started talking with Microsoft three and a half months ago, and two weeks after, we started coding. Three months down the road, we are already going live," shares Colegrave. The collaboration with Microsoft provided the technological foundation required for such swift progress.
Azure OpenAI Service
Syensqo relied on the seamless integration of Azure AI Services for the realization of their use case. "We leveraged Azure AI Services and the general Microsoft Azure technology stack, Azure OpenAI Service in specific." This foundation ensured that Syensqo could deploy SyGPT at scale, meeting the demands of their global operations.
Syensqo.AI also leverages Azure Cosmos DB for globally distributed, multi-model database management, ensuring high availability and low latency. The integration of Azure API Management ensures secure and scalable API connections to existing databases, apps, and other data storage services within Syensqo, maintaining data safety and privacy.
We started talking with Microsoft three and a half months ago, and two weeks after, we started coding. Three months down the road, we are already going live.
Vincent Colegrave, Head of Syensqo.AI, Syensqo.AI
Internal existing datasets as a source
The sources of AI for SyGPT will come from API connections to existing databases, applications, and other data storage services within Syensqo. These API connections are secured to ensure the safety and privacy of the data. For example, Syensqo’s extensive research and development database, which contains years of scientific research, experimental results, and proprietary data, will be integrated via Microsoft Azure API Management. This will allow SyGPT to utilize a vast amount of internal data, making it a powerful tool for research and data analysis.
SyGPT aims to have a positive impact within the organization
SyGPT benefits a wide range of stakeholders within Syensqo, enhancing data access, operational efficiency, and collaboration through a centralized, secure data hub powered by AI.
The impact of SyGPT has been significant. "We tested SyGPT with 80 colleagues who spent almost a month querying and giving feedback. We reported and analyzed more than 100 bugs," states Colegrave. The feedback has been overwhelmingly positive, highlighting the efficiency and security of the tool.
This rigorous testing phase not only validated the tool's capabilities but also provided valuable insights for future improvements. Vincent Colegrave, Head of Syensqo.AI elaborates, "We want to embed more models and follow the fast-evolving technology. Our goal is to bring the best value to our colleagues."
Partnership
The partnership between Syensqo.AI and Microsoft has led to the successful development and deployment of SyGPT, a chatbot that significantly enhances data access and operational efficiency. With a strong technological foundation and a shared commitment to innovation, both companies are poised to explore further advancements in AI technology.
We want to embed more models and follow the fast-evolving technology. Our goal is to bring the best value to our colleagues.
Vincent Colegrave, Head of Syensqo.AI, Syensqo.AI
Follow Microsoft