Enterprise Generative AI Framework: Establishing AI Excellence at Industry Scale

A major US airline engaged Taller through a staffing partner in March 2023, with work spanning applied AI and core modernization. Taller ran six pods across AI/data and .NET/Angular rebuilds, with multi-year extensions confirmed.

The airline’s ambition was to become an applied-AI leader in transportation. That required a single framework any internal team could build against, with the operational discipline (governance, monitoring, and evaluation) that turned experimentation into deployable capability.

Taller helped design and develop the reusable generative-AI framework: a foundation for building AI agents, orchestrating multi-agent workflows, and running graph-based retrieval, usable across many internal projects. What made it load-bearing rather than aspirational was treating each agent as a first-class unit of deployment: every agent had its own versioned definition, its own scoped knowledge base, its own defined actions, and its own evaluation checks. The framework supplied the orchestration spine (LangGraph), the retrieval layer (graph-based RAG with re-ranking), a model-provider abstraction (AWS Bedrock, plus other model routes when cost or speed demanded), and the knowledge-graph layer that let agents share structured context without leaking one domain’s language into another.