Field Research for the 2nd Edition
Rethinking Capital Project Delivery: Agentic AI-Driven Strategies for the New Era
I’m collecting real examples—what changed, what stuck, what didn’t—as engineering timelines compress and requirements change accelerates. The aim is to capture practical lessons that delivery leaders can use immediately—without hype.
No project names required. Anonymous contributions welcome.
What I’m looking for
Real project stories where AI or workflow changes actually moved time, risk, or rework.
I'm especially interested in real project stories where teams have measurably accelerated end-to-end delivery by combining AI and new ways of working.
A good benchmark for the kind of “speed to first value” I mean is xAI's Colossus — built in 122 days. That's radically different from the typical 12–24+ month cycle most large builds work to. The gap isn't magic tech — it reflects a different engineering mindset.
Removed waiting: Where AI (or AI-enabled workflows) removed waiting and sped up delivery
Absorbed late change: Where it helped teams absorb late change without derailing execution
Decisions & control: What it improved in decision management / change control / approvals
Interfaces & dependencies: What it changed in interfaces & dependencies (TQs/RFIs, tie-ins, handoffs)
Controls & performance: What it changed in project controls & performance management (faster recovery action, fewer surprises)
Stage gates: How it affected stage gates (pack readiness, assurance speed, evidence)
Safety & risk assurance: How it affected safety and risk assurance as plans changed (HAZID/HAZOP/SIMOPS becoming “living”)
How to contribute (choose what’s easiest)
30–45 min interview (best for capturing detail)
10–15 min written input (quick structured response)
Anonymous contribution (if your organisation requires it)
Confidentiality
I understand most project examples come with sensitivities (NDAs, commercial positions, HSSE, internal governance). You can share at a level that’s safe for you and your organisation.
- No project names required
- Sensitive details can be removed/redacted
- Attribution only with explicit permission
- If your contribution is included in the 2nd edition, we’ll confirm the final wording and attribution with you first.
Research for the 2nd Edition: Share What Worked
Use the form below to share one change that actually pulled time, risk, or rework out of delivery.