doctrine.work

About

Doctrine emerged from repeatedly encountering the same organizational questions in very different environments.

Over the past decade, I've worked across fintech, enterprise AI, climate technology, conversational systems, AI infrastructure, and edtech.

Different industries. Different products. Different tech.

And yet the same organizational patterns kept preventing the technology from producing the change it promised.

Why do capable systems fail to produce transformation?

How do incentives shape adoption and resistance?

What remains fundamentally human as AI becomes increasingly capable?

Doctrine emerged from trying to answer those questions — first for myself, then for the organizations I work with.

How the Perspective Formed

I didn't start by studying organizational transformation.

I was building AI products and repeatedly encountered the same structural constraints in very different contexts. Each organization revealed a different part of the same larger system.

Over time, those observations stopped feeling industry-specific and started feeling fundamental. What began as pattern recognition became a way of understanding organizational change.

Doctrine is an attempt to codify that understanding.

Where the Framework Came From

Razorpay
Fintech · Risk Systems
Accuracy alone isn't enough. Explainability and trust are integral to the product.
Building risk systems taught me that probabilistic models need deterministic guardrails and explainability, particularly when decisions carry institutional consequences. A model can be 99% accurate and still be undeployable if no one can explain a single decision.
→ The Trust Budget
What Doctrine learned
Trust is built through explainability, not accuracy alone.
AiDash
Climate Tech · AI Platforms
What changes when humans and AI share responsibility.
I led the transition from separate AI and human workflows to an integrated human-in-the-loop platform. It showed me that improving the model and improving the system are different problems.
→ The Agentic Transition
What Doctrine learned
Scaling AI is fundamentally an organizational problem.
iMerit
AI Infrastructure · Data Operations
Humans aren't a fallback. They're the liability owners.
Scaling data operations taught me that humans in the loop aren't just a safety net for AI — they are the source of ground truth and the owners of accountability. Human responsibility is a constraint that cannot simply be delegated.
→ The Human Moat
What Doctrine learned
Human judgment is an architectural component, not a workaround.
Wyzion
Building systems that turn conversational intelligence into organizational action.
Wyzion is where these ideas are being applied in production.
The work involves more than the intelligence layer. It requires connecting conversations, organizational context, decision systems, and human intervention into a single operating system.
What Doctrine is continuing to explore
How organizations can build systems that improve judgment rather than simply automate work.

Recurring Patterns

01
The technology is usually the easy part.
The difficult questions are almost always about incentives, governance, and operating models.
02
Most complexity starts as a reasonable local decision.
Over time, those decisions accumulate into systems no one intended to build.
03
Avoided decisions become the most expensive ones.
They become more political, harder to unwind, and more costly the longer they wait.

Why Doctrine Exists

Doctrine is less a consulting practice than an attempt to codify how organizational transformation actually happens — and why it so often doesn't.

The writing codifies the ideas.

The methodology applies them during engagements.

The case studies test them against organizational reality and provide the feedback loop.

The advisory practice helps organizations navigate them.

Today

Today, Doctrine exists as a connected system.

The Doctrine
The ideas.
Methodology
The reasoning process.
Case Studies
The ideas under real constraints.
Advisory
The practical application.

Each exists to help organizations understand what must fundamentally change before deciding how to change.