What We Think · Podcast
Elixir AI Dispatch. Conversations with the practitioners building and operating enterprise AI in production. Not panel discussions. Not predictions. Decisions and outcomes from real deployments.
Every episode of Elixir AI Dispatch features a conversation with an Elixir practitioner, a client stakeholder, or a domain expert who has made a specific decision about enterprise AI in production. What they decided. What happened. What they would do differently. Available on all major podcast platforms.
Recent episodes
Conversations from the practitioners operating enterprise AI.
01/ 07
Finance leaders on what changed when AI ran the close: Three CFO conversations from 2026
Three CFOs share the same observation: the month-end sprint did not disappear when AI arrived. The remaining work is the hardest to automate. What that means for finance teams planning AI deployments in 2026.
What Codelco showed us about sourcing automation at scale: 250 RFP events per day, zero buyers involved
The Kyyte delivery lead on the Codelco SourcifAI engagement. What made it work, what the 500+ suppliers per RFP event required from the data architecture, and what 6-8 FTE saved actually looks like in a procurement team.
The data foundation problem: Why 60% of enterprise AI programmes stall at the same point
Elixir's AI data practice lead on what the AI Outlook 2026 research found about data foundation failures. The specific types of data quality issues that stop an AI build mid-programme, and how the Agentic AI Framework's Layer 1 addresses them before the build begins.
Claim denied: How SAP MedTech AI reduced a Gulf healthcare group's denial cycle from 14 days to under 9
The Healthcare RCM Hypercase in full. The delivery lead on the NLP model selection, the governance requirements for a Gulf healthcare environment, and why the outcome is still being improved twelve months after go-live.
Permanent Operations: Why most AI programmes fail after go-live and what Elixir does differently
Elixir's Agentic Operations lead on Layer 5 of the Agentic AI Framework. What Permanent Operations actually involves day to day, how 314 agents are currently being operated, and what the monthly outcome report looks like.
VPRC and the patent behind the agent: How owned research reaches production inside SAP
VPRC's research director on the path from the multi-agent orchestration patent to the coordinated agent architecture inside SourcifAI and the six Kyyte products. What research-to-production actually looks like on a 12 to 18 month cycle.
The AI-First Manifesto: Five principles and why each one came from a programme that failed without it
Elixir's practice leads on the origin of each of the five AI-First Manifesto principles. Each principle is traced back to a specific programme failure pattern that the principle was designed to prevent. The AI Outlook 2026 research findings presented in context.
Listen on your platform
Elixir AI Dispatch is available on all major podcast platforms.
Spotify
Search Elixir AI Dispatch
Apple Podcasts
Search Elixir AI Dispatch
Google Podcasts
Search Elixir AI Dispatch
YouTube
Elixir Global channel
Conversations with the practitioners who built the systems being discussed.
Subscribe to Elixir AI Dispatch on your platform of choice. New episodes monthly. Every episode cites the production deployment it is grounded in.
Recent episodes
Ep 07: Finance leaders and the closeEp 06: Codelco and SourcifAIEp 05: The data foundation problemEp 04: Healthcare RCM HypercaseTopics covered
Agentic AI and Permanent OperationsSAP and procurement AIHealthcare RCMFinance operationsAvailable on
SpotifyApple PodcastsGoogle PodcastsYouTube (@elixir_glob)Related
Perspectives BlogWebinars and Live SessionsNewsletter: The Signal