# AI Orchestration for Smart Cities and the Enterprise with Robin Braun and Luke Norris - #755

Page: https://stenobird.com/podcast/twiml-ai-podcast/ai-orchestration-for-smart-cities-and-the-enterprise-with-robin-braun-and-luke-norris-755
Text version: https://stenobird.com/podcast/twiml-ai-podcast/ai-orchestration-for-smart-cities-and-the-enterprise-with-robin-braun-and-luke-norris-755.md
Podcast: [The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)](https://stenobird.com/podcast/twiml-ai-podcast)
Published: 2025-11-12T20:05:00+00:00
Episode link: https://twimlai.com/podcast/twimlai/ai-orchestration-for-smart-cities-and-the-enterprise/
Audio file: https://pscrb.fm/rss/p/traffic.megaphone.fm/MLN7553368301.mp3?updated=1762978440
Processing state: processed
JSON: https://stenobird.com/v1/public/podcasts/twiml-ai-podcast/episodes/ai-orchestration-for-smart-cities-and-the-enterprise-with-robin-braun-and-luke-norris-755
Duration seconds: 3286

## Resource

AI orchestration is moving from experimental mandates to tangible ROI by automating complex enterprise workflows and legacy data extraction. Using the 'Agentic Smart City' in Vail, Colorado, as a blueprint, the discussion demonstrates how agentic workflows can solve specific, high-impact problems like accessibility compliance and fire risk assessment.

## Highlights
- Main idea: The focus of enterprise AI has shifted from broad mandates to proving measurable ROI through targeted use cases
- Practical takeaway: Use a 'mud puddle by mud puddle' approach—solving small, discrete problems like 508 compliance before attempting massive system overhauls
- Failure mode: Attempting grandiose, all-encompassing AI visions at the start can lead to project collapse under the weight of complexity
- Technical insight: Modern AI agents can use 'computer use' capabilities to navigate browsers and remediate web accessibility issues automatically
- Strategic advice: Do not wait for perfect data hygiene; the effort to cleanse legacy data often results in data that is already obsolete by the time it is ready

## Topics

AI Orchestration, Agentic Workflows, Smart Cities, Enterprise AI, 508 Compliance, Data Engineering, Hybrid Cloud, Computer Use Agents

## Chapters
- 1:00 — The Shift from AI Mandates to ROI: Discussion on how the enterprise focus has moved from general AI adoption to seeking tangible returns on investment.
- 5:05 — Automating Ontology and Entity Extraction: How recent advancements in models allow for automatic creation of ontology schemes and abstraction from legacy files.
- 9:10 — Automating 508 Compliance: A deep dive into using visual language models and agents to automate the manual, error-prone process of web accessibility remediation.
- 17:10 — AI for Environmental Risk Assessment: Using contextual data and camera feeds to predict and manage fire risks in smart city environments.
- 21:25 — Leveraging Existing Infrastructure: The importance of reusing existing hardware, such as municipal camera networks, to deploy AI solutions cost-effectively.
- 33:35 — The Mechanics of Agentic RAG: An explanation of the recursive lookup, ontology systems, and vector representations required for advanced agentic retrieval.
- 46:05 — Managing the AI Stack and Lifecycle: Addressing the complexities of maintaining AI infrastructure and the importance of managing the lifespan of data assets.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/twiml-ai-podcast/episodes/ai-orchestration-for-smart-cities-and-the-enterprise-with-robin-braun-and-luke-norris-755/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/twiml-ai-podcast/ai-orchestration-for-smart-cities-and-the-enterprise-with-robin-braun-and-luke-norris-755.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.