Mash

Mash is developing AI-powered knowledge agents that help B2B revenue teams access accurate information across messaging platforms, CRMs, project management tools, and internal documentation. As organizations generate increasing volumes of fragmented data, a major technical challenge is enabling AI systems to connect information across multiple sources while understanding the relationships between people, projects, customers, and business activities.

TICAN is enabling Mash to hire two graduate students to work on advancing applied AI research projects focused on improving how these systems retrieve, connect, and evaluate information across complex enterprise environments. One project explores graph-based retrieval methods that enable relationship-aware reasoning across multiple data sources, while the other is developing a continuous evaluation and optimization framework to improve AI reliability and performance over time.

Together, the projects address a growing challenge in enterprise AI: ensuring large language model (LLM) systems can deliver accurate, scalable, and trustworthy outputs in real-world business settings.