In the past weeks, we established the Handshake Protocol and the Target Architecture. We know the rules of the network, and we know our destination.

But a system only runs efficiently if you distribute the compute properly. One of my core jobs as an engineering leader is Opportunity Mapping.

I keep a mental map of where every single person on my team (directs and skip-levels) currently stands. Every individual contributor has their own unique topology, their own way of interacting with the world and their own specific way of being amazing.

Although it can be a lot more nuanced than that, I tend to map engineers along a spectrum between two distinct hardware profiles:

The Two Hardware Profiles

The Vanguards

The headfirst explorers. They excel at discovery, testing the limits of the system, and moving with extreme velocity. They aren't afraid to break things to find the path.

The Sentinels

The structured methodologists. They are cautious, organized, and look around every corner to ensure nothing goes wrong before execution. They build the paved roads.

If you treat these profiles like generic, interchangeable resources, you will crash your system.

Here is how I actively load-balance the organism:

The Load Balancing Protocol

1. Cross-Wiring (Pairing Protocol)

I frequently pair a Vanguard with a Sentinel. The Vanguard learns to temper their chaos and think about edge cases. The Sentinel learns to tolerate risk and increase their execution velocity. By exposing them to their opposite, I force the system to fine-tune itself.

2. Contextual Deployment

Every project has its own topology. If I'm running a fast, pilot LLM/AI experiment where we need to discover limits fast, I deploy a squad of Vanguards. But if the project is core Anti-Money Laundering infrastructure where a single bug is catastrophic? I deploy the Sentinels. You match the risk profile of the node to the risk profile of the work.

3. The Alternating Current

No one should feel like unused compute power, but no one should constantly overheat, either. I alternate their workload: one project inside their natural comfort zone (to build confidence and high-impact momentum), followed immediately by a stretch project that pulls them completely out of their element (to force adaptation).

The Leadership Reality Check

Engineers are not config files you can just copy-paste from one Jira epic to another.

If someone on your team is underperforming, there is a high probability they aren't failing โ€” you just deployed a Sentinel to do a Vanguard's job, or vice versa.

Opportunity mapping is how you ensure everyone plays a game they can actually win, while the system maintains a constant, predictable rhythm of delivery.


Read it on LinkedIn: Field Note #12 on LinkedIn