Once agents can use tools, ordinary business content can become part of the control surface. Documents, tickets, webpages, records, and retrieval results may contain instructions the agent should read as data, not follow as commands. Based on a conversation with Pramod Krishnan from PwC, this piece looks at indirect prompt injection, tool permissions, trace review, and why production agents need a clear separation between content and action.

Pramod Krishnan is a Managing Director - AI Managed Services at PwC, specializing in enterprise AI transformation — helping large organizations move from AI experimentation to production operating models. In this episode with Demetrios, Pramod breaks down exactly what the OpenClaw wave means for enterprises, and the control frameworks PwC uses before a single agent touches production.

Production agent ROI is usually calculated too narrowly. The model bill is visible, but the bigger cost often sits in reviewer time, eval maintenance, retrieval, storage, platform work, and process change. Based on a conversation with Rani Radhakrishnan from PwC, this piece argues that the real comparison is not headcount saved minus inference spend, but the full cost of the old process against the full cost of the agent-assisted one.

MLOps Community is joining the Linux Foundation as the official user group of the Agentic AI Foundation. The community continues, with more support behind the events, newsletter, podcast, and practitioner conversations.

Before MCP was a standard and before LangChain was widely adopted, his team had already shipped their own orchestration layer and tool protocol in production. This conversation is a rare look at what it takes to build an agentic system that actually books trips, runs on WhatsApp, and keeps adding capabilities without falling over.

Subtitle: It’s a feature of the architecture Summary: Hallucination in LLMs is not a data quality problem. It is not a training problem. It is not a problem you can solve with more [RLHF](https://en.wikipedia.org/wiki/Reinforcement_learning_from_human_feedbac), better filtering, or a larger context window. **It is a structural property of what these systems are optimized to do.** I have held this position for months, and the reaction is predictable: researchers working on retrieval augmentation, fine-tuning pipelines, and alignment techniques would prefer a more optimistic framing. I understand why. What has been missing from this argument is geometry. Intuition about objectives and architecture is necessary but not sufficient. We need to open the model and look at what is actually happening inside when a system produces a confident wrong answer. Not at the logits. Not at the attention patterns. At the internal trajectory of the representation itself, layer by layer, from input to output. That is what the work I am presenting here did.

Anurag Beniwal (Member of Technical Staff at ElevenLabs) breaks down the real-world challenges of building voice agents—from latency, transcription accuracy, and turn-taking to the tradeoffs between cascaded systems and end-to-end speech models. The conversation explores why production systems rely on “constellations” of models, how to design for non-technical users (especially in customer support), and why voice unlocks richer context—but introduces far more complexity than chat. Ultimately, it’s a deep dive into making voice AI practical, reliable, and usable at scale.

A deep dive into the practical limitations of agent protocols like MCP and A2A for low-level tasks, and why the "Linux philosophy" of using a raw command-line interface provides a more lightweight, composable alternative for local development, paving the way for an Agent OS.