The Engineering of AI Agents: Context, Harnessing, and Autonomy

Name: The Engineering of AI Agents: Context, Harnessing, and Autonomy
Uploaded: 2026-05-07T04:09:45Z
Duration: 42 min 2 s
Channel: InfoQ
Description: Birgitta Böckeler explores the shift from prompt engineering to harness engineering, building deterministic safety nets around non-deterministic AI models.

InfoQ

234K subscribers • 59 videos on Hallway • Conference channel

YouTube

One of the most-watched uploads on InfoQ, a conference or university channel Hallway lists. Nobody has recommended it in the communities yet, so it ranks below community picks.
Presented by Birgitta Böckeler, Thoughtworks
May be out of date. The tools, versions or practices it shows have likely moved on.

Intermediate. Assumes working knowledge.

27K views5 months ago

AI and machine learning Software design AI agents and tool use Transformers and LLMs Software architecture

Birgitta Böckeler explores the shift from prompt engineering to harness engineering, building deterministic safety nets around non-deterministic AI models.

📩 Subscribe to the InfoQ Weekly newsletter! No hype. No fluff. Just the signals senior engineers actually care about - from #AI, #DevOps, and #Java to #CloudComputing & #SoftwareArchitecture. 🗓️ Delivered every Tuesday, it's your quick round-up of innovator & early-adopter technologies. Join 250,000+ developers who read it to stay ahead 👉 infoq.com/news/#infoq-nl ************************************************ As we move from simple autocompletion to autonomous AI agents, "vibe coding" isn't enough for enterprise-grade software. Birgitta Böckeler (Thoughtworks) explores the shift from prompt engineering to "Harness Engineering"—the practice of building deterministic safety nets around non-deterministic models. ⏱️ Video Timestamps (For Navigation) – The Evolution: From Autocomplete to Agents – What is Context Engineering? (Rules, MCP, and Skills) – Modularizing Context with Anthropic Skills – The Cost Reality: $380/day per Developer? – Subagents and Research Workflows – The "Lethal Trifecta" of AI Security Risks – Harness Engineering: Building Deterministic Safety Nets – Maintaining Quality: Structural Tests as Feedback Loops – The "Goldilocks Speed": How much velocity do we actually need? – Final Thoughts: Scaling AI Autonomy Safely 🔗 Transcript available on InfoQ: bit.ly/4we3odZ #SoftwareArchitecture #AIAgents #GenerativeAI #SoftwareEngineering #ClaudeCode #GitHubCopilot #ContextEngineering

Why it's here

A top upload from a listed channel

InfoQ is on Hallway's list of conference and university channels, and this upload is one of its most-watched talks or lectures. Nobody has recommended it in the communities yet, so it ranks below community picks.