Xavi Ablaza

CEO of cloud infrastructure portfolio. 7 years operating.

P&L engineering × technical leverage

Every operating decision is a P&L decision.

Capital allocation is an engineering problem. Being a P&L leader involves the discipline of turning technical work into margin, velocity, resilience, and option value. Here are the playbooks you need to produce the alpha you need to make returns above your operating baseline.

P&Ltechnical decisions expressed as margin, cash, risk, and option value

5 loopsmap, price, allocate, ship, compound — measured together

1 graphcustomers, product, systems, teams, spend, and ownership connected

The System

Map→ Price→ Allocate→ Ship→ Compound

Five steps, applied to every platform, team, process, and reliability decision. The loop does not stop after the migration ships.

The Operator

Allocator· Operator· Builder· Scientist

Four lenses, each common alone — the combination is what compounds.

Build the P&L engineering substrate

I work on organizations where engineering is no longer just delivery capacity: it is the control plane for margin, reliability, cost, trust, velocity, and organizational learning.

Email GitHub LinkedIn

Your business is a directed graph →

Your P&L is a graph of nodes and edges. Every cost center, revenue line, and causal link between them. Walk it and you find the mispriced edges where value is leaking. Three frameworks show you how.

Price the opportunity in dollars, not vibes →

Every operating opportunity has a return distribution, an error cost profile, and a Sharpe ratio. You just haven't calculated them yet. Five interactive tools do the math before you write the code.

Construct the portfolio →

Rank your instruments by risk-adjusted return, mapping your tolerances & handling correlations between bets. This is the same efficient frontier math that your CFO uses for physical capital, applied to the CTO's roadmap.

Systems that converge by construction →

The goal is convergence by construction. Design the system so improvement is a mathematical property, not an aspiration. Ratcheted quality gates and graduated autonomy compress execution risk into something you can manage.

Build the appreciating asset →

Models depreciate through distribution shift. Verifiers, labeled corpora, and institutional rubrics appreciate through operating use, one of the only capital asset classes that does. The framework tells you which side deserves your next dollar.

Frameworks

The Factory Typology →Knowledge work is manufacturing — but which factory? Name the input, process, and output signature, then borrow the mature operating playbook. Your Chart of Accounts Is a Directed Graph →Every cost and revenue line is a node. Every causal relationship is an edge. Walk the graph and find the mispriced edges where value leaks. The Performance Frontier →Locate where excellence lives when nobody agrees what good looks like: map the distribution, find the 99th percentile, compute the gradient. The Demand Field →Demand is a force field on the optimization landscape. It is fixed, hidden, and acts on the product whether you measure it or not. Capital Allocation →Every operating decision is a capital allocation decision. Rank instruments by risk-adjusted return, map tolerances, and construct the portfolio. Kill Protocol →The allocator skill is saying no: marginal Sharpe, left-tail survivability, and base-rate verification before a bet consumes the slot. The Operating Efficient Frontier →The set of portfolios that maximize expected return at each risk level. Points below the frontier are dominated; the question is which point you choose. The Risk Tolerance Map →Map cash, runway, covenants, and narrative constraints into the preferred point on the efficient frontier. Different firms can correctly choose different points. Correlated Execution Risk →Operating portfolios are not sums of individual NPVs. Shared teams, stacks, vendors, and review queues create shared failure modes. Designed Convergence →Design the game so convergence to the desired outcome is structural, not aspirational. Mechanism design meets Bayesian ratchet search. The Deity Problem →Your AI is trying to serve you but cannot read your mind. Use structured elicitation, revealed preference, and direct query to learn what the operator wants. Quality Hillclimb →Apply ratcheted quality gates to stochastic agent output. The agent does not need a plan — the gates create ascent. The Promotion Protocol →A three-state progression for safely giving AI more independence. Promote on demonstrated performance; roll back on drift. Knowledge Work as Capital →Every piece of knowledge work either compounds or depreciates. Models decay; data, verifiers, and rubrics can appreciate through use. The Capital Value of Verifiers →The verifier is one of the rare capital assets that appreciates through operating use: every failure it catches gets encoded into the next run.

Concept Graphs

Curated

Frameworks →Nine-plus strategic models for finding, pricing, and allocating operating value. Tools →Interactive calculators for pricing operating instruments — diagnose, calibrate, invest, value, and screen. Lexicon →Defined terms invented to name operating realities that existed but had no shared word. Positions →Four overlapping circles — Allocator, Operator, Builder, Scientist — and the intersections where the value lives. Thesis →The full lattice: every framework, tool, and term organized by allocation stage. I Have This Problem →Problem-first entry points into the frameworks, tools, and operating moves. Knowledge Graph →Interactive map of how frameworks, tools, terms, positions, and posts connect.

Substrates

Business →The operating substrate: markets, teams, incentives, pricing, and organizational control loops. Money →Capital decisions structured as a dependency tree — the why behind every allocation move. Math + ML →Concepts from arithmetic to transformers, ordered by dependency with interactive visualizations.

Meta

About →The map behind the operator, allocator, builder, and scientist positions. Experiments →Working notes and trials around the system.