peopleanalyst

← Portfolio

AnyComp.AIanycomp.ai

AnyComp.AI — the compensation decision OS, run on your data; white-glove, not a tool you assemble.

AnyComp.AI — the compensation decision OS, run on your data; white-glove, not a tool you assemble.

The compensation concierge — comprehensive comp analytics plus white-glove service for executives. AnyComp.AI is the org-wide Compensation Decision OS you graduate to from the self-serve Compensation Toolbox: pay decided across all your jobs and people at once, on a replayable cycle audit. The comp parallel to Performix — a polished concierge product off the store.

The problem

Compensation can't honestly be done one job at a time: ranges, equity, and budget interact, so a single-role view is always incomplete. Yet merit lives in spreadsheets that fork every cycle — formulas vanish, approvals leave no record, and 'what changed vs plan' becomes forensic archaeology. At the executive tier the need isn't another self-serve tool; it's a decision system run with white-glove rigor across the entire org.

What I built

Live at anycomp.ai: the compensation decision OS, run on your data. AnyComp.AI turns executive intent into optimized, budget-constrained compensation actions — comprehensive analytics across your whole org, run for you, white-glove, not a tool you assemble. Underneath: versioned comp models (levels, geography, midpoint curves), market-anchored band math against OEWS/SOC labor data, a field-leading multiple-regression pay model with a factor decomposition and honest error bars, merit / variable-pay / equity / discretionary engines, and a strategy → priorities → objective → optimizer → simulator → scenarios decision loop that returns several scored options ('never one option') on a replayable cycle-audit trail.

What's novel
  • 01The honest model, not a number. The market-data industry sells a static percentile cut dressed as truth; AnyComp.AI sells the model + the decomposition + the error bars — pay as a factor-decomposed, uncertainty-quantified distribution. 'There is no single correct market value — and we prove it.' Honest uncertainty is the feature that makes the answer defensible in front of a board.
  • 02Comp as a holistic decision, not a job-by-job spreadsheet. The Decision OS runs scenarios across all jobs and people at once, because ranges, equity, and budget interact — the thing a single-role view structurally can't do.
  • 03Decisions, never one option. The loop returns several scored scenarios across multiple value measures with a full audit trail — comp as a versioned, replayable service, not a lost workbook tab.
  • 04Concierge above a store, the cross-over tier. AnyComp.AI is to Compensation Toolbox what Performix is to the People Analytics Toolbox — the executive-oriented, run-it-for-you product that categorically crosses over the deconstructed professional toolbox: white-glove delivery on your data, not a kit you assemble.
Outcome

Live at anycomp.ai — the executive-tier compensation concierge, sitting above the self-serve Compensation Toolbox the way Performix sits above the People Analytics Toolbox. Strong monetization characteristics on the executive tier; the same market-anchored substrate also feeds a public job-and-pay browser and a personalized 'grow your pay' pathway in active build.

AnyComp started from a real truth: you can't set pay job-by-job, because ranges, equity, and budget all interact — so it became a Compensation Decision OS that runs scenarios across the whole org. The self-serve Compensation Toolbox fronts it for comp pros; AnyComp.AI is the concierge above it — the executive, white-glove face where the analysis is run for you across every job and person at once. It crosses over the deconstructed professional toolbox the way Performix does: same engine underneath, but pitched at the executive who wants the decision, not the kit. One model, two doors: pick a capability off the shelf, or hand the whole org's pay decision to the concierge.