Optimize adaptive analytics plan

What it's for

Lets a Gitrevio agent decide what to learn next—and when to stop analyzing—by building an auditable outcome-contingent evidence plan instead of calling every tool or following a fixed checklist.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "actions": [
    {
      "id": "build",
      "loss_by_state": [
        100,
        0
      ]
    },
    {
      "id": "defer",
      "loss_by_state": [
        0,
        40
      ]
    }
  ],
  "analyses": [
    {
      "cost": 1,
      "dependency_ids": [],
      "duration_periods": 0,
      "exclusion_ids": [],
      "id": "prototype",
      "results": [
        {
          "id": "negative",
          "likelihood_by_state": [
            0.9,
            0.1
          ]
        },
        {
          "id": "positive",
          "likelihood_by_state": [
            0.1,
            0.9
          ]
        }
      ]
    }
  ],
  "analytics_budget": 1,
  "maximum_analysis_steps": 1,

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "configuration": {
    "action_count": 2,
    "analysis_count": 1,
    "analytics_budget": 1,
    "cost_of_delay_per_period": 0,
    "maximum_analysis_steps": 1,
    "maximum_exact_histories": 250000,
    "maximum_total_duration_periods": 10,
    "minimum_expected_net_value": 0,
    "state_count": 2
  },
  "decision": "run_adaptive_analytics_plan",
  "guardrails": [
    "State priors, action losses and every analysis likelihood require prospective local calibration under one decision perimeter; an exact solver cannot repair omitted states, biased tests or invalid likelihoods.",
    "Analysis results are conditionally independent given the represented state. Correlated measurement errors, shared data leakage or result-dependent costs require an expanded observation model.",
    "The policy may stop early and is optimal only inside its bounded action, analysis, dependency, budget, delay and horizon model. Safety, legal, fiduciary and implementation authority remain external constraints."
  ],
  "method": "exact_adaptive_bayesian_analytics_planning_v1",
  "policy_tree": {
    "act_now_expected_total_loss": 20,
    "analysis_cost": 1,
    "analysis_id": "prototype",
    "branches": [
      {
        "next": {
          "action_id": "defer",
          "decision": "act",
          "posterior_expected_total_loss": 4,
          "posterior_state_probabilities": {
            "adverse": 0.9,
            "favorable": 0.1
          }
        },
        "predictive_probability": 0.5,
        "result_id": "negative"
      },
      {
        "next": {
          "action_id": "build",
          "decision": "act",
          "posterior_expected_total_loss": 10,
          "posterior_state_probabilities": {
            "adverse": 0.1,

How it works

Sequential Bayesian & bandits — Learn while deciding — update beliefs as evidence arrives and choose where the next unit of effort is worth spending.

1. 1 Freeze mutually exclusive latent states, prior probabilities, terminal action losses and candidate analyses with discrete result likelihoods, cost, duration, dependencies and exclusions.
2. 2 At every posterior belief compare acting now with every feasible next analysis, update state beliefs by Bayes' rule for each possible result, and solve the bounded policy tree exactly with memoized dynamic programming.
3. 3 Return the first analysis and full contingent policy only when net value clears the governed gate; reconcile act-now loss, best fixed sequence, adaptivity value and perfect-information upper bound with a global certificate.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "choose an exact adaptive sequence of" }
  → finds "optimize_adaptive_analytics_plan"

gitrevio_capability_describe
  { "capability_id": "optimize_adaptive_analytics_plan" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "optimize_adaptive_analytics_plan", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in Analytics assurance & orchestration →