Last October, IRIS by Argon & Co launched EDGAr, an internal AI chatbot, with a single mission: help all Argon & Co consultants leverage the company internal knowledge management platform to search for, discover and generate new contents. To go further, the IRIS by Argon & Co team was looking to design a solution that helps make AI truly accessible and relevant in the company day-to-day tasks. A solution that would be:<\/p>\n
\u2022 Secure:<\/strong> GDPR compliant, data not used to train LLMs, data encrypted and stored in the EU These requirements are probably the same for every knowledge worker and yet, no solution could answer them all. So they built one. Today, the team is proud to reveal \u201cEDGAr Assistant Builder\u201d: a no-code, self-service platform to build and share custom AI assistants, based on Microsoft Azure architecture. This platform is already used by Argon & Co consultants, to support daily tasks and by IRIS by Argon & Co data scientist, to quickly PoC a use-case by our clients. Now, the teams are able to start helping clients of all sizes to achieve the same objective: embracing the AI transformation !<\/p>\n
\n\u2022 Knowledge-based:<\/strong> selection of documents by the user, OCR, citation of exact sources\u2026
\n\u2022 Configurable:<\/strong> choice of LLM, raw context or leveraging RAG, customizable system prompts\u2026
\n\u2022 Self-service:<\/strong> no-code approach, accessible to anyone with basic training on GenAI
\n\u2022 Extendable<\/strong>: with capabilities to be added progressively in an open platform environment<\/p>\n