Algorithm Catalog
17 production-grade algorithms across 8 categories. All endpoints return results in under 25ms with zero LLM cost.
Optimize
Evolutionary and bandit algorithms for finding optimal solutions in large search spaces.
Multi-Armed Bandit
$0.01UCB1, Thompson Sampling, and Epsilon-Greedy algorithms for explore/exploit decisions. Maximizes cumulative reward across options with unknown payoffs.
Contextual Bandit (LinUCB)
$0.02Context-aware decision making using linear upper confidence bound. Uses feature vectors (time, energy, urgency) to make personalized selections.
Genetic Algorithm
$0.10Multi-objective evolutionary optimization with Pareto frontier. Supports tournament, roulette, and rank selection with configurable crossover and mutation.
CMA-ES
$0.10Covariance Matrix Adaptation Evolution Strategy for continuous black-box optimization. State-of-the-art for derivative-free numerical optimization.
Simulate
Monte Carlo and scenario simulations for risk assessment and what-if analysis.
Monte Carlo Simulation
$0.05Run thousands of probabilistic simulations with configurable distributions. Supports normal, lognormal, uniform, triangular, beta, and exponential.
Scenario Planning
$0.08What-if analysis comparing multiple scenarios against a base case. Calculates outcome deltas, per-variable impact, and sensitivity rankings.
Solve
Constraint satisfaction and pathfinding for scheduling, routing, and resource allocation.
Constraint Optimizer (LP/MIP)
$0.10Linear and mixed-integer programming using HiGHS solver (WASM). Production-grade LP/MIP/QP solver for resource allocation and scheduling.
Schedule Optimizer
$0.10Optimal task-to-slot assignment using constraint satisfaction. Respects deadlines, priorities, and resource constraints.
Analyze
Graph analysis, portfolio risk, and multi-model consensus for complex data interpretation.
Decision Graph
$0.05Graph analysis with PageRank, Louvain community detection, and shortest path using graphology. Model decisions as interconnected nodes.
Portfolio Risk (VaR/CVaR)
$0.10Value at Risk and Conditional VaR for portfolio risk assessment. Computes expected return, volatility, and tail risk metrics.
Predict
Bayesian inference, ensemble methods, and time series forecasting for probabilistic predictions.
Bayesian Inference
$0.02Bayesian belief updating with configurable priors and evidence factors. Computes posterior probability, calibration score, and credible intervals.
Ensemble Model
$0.05Combine multiple model predictions using weighted voting, stacking, or Bayesian model averaging. Quantifies epistemic and aleatoric uncertainty.
Time Series Forecast
$0.05ARIMA and Holt-Winters forecasting for time series data. Produces point forecasts with confidence intervals and model diagnostics.
Detect
Statistical anomaly detection and signal scanning for identifying unusual patterns.
Anomaly Detection
$0.02Statistical anomaly detection using Z-Score and IQR methods. Identifies outliers in numerical data with configurable sensitivity thresholds.
Score
Convergence and calibration scoring for evaluating prediction quality and source agreement.
Convergence Scoring
$0.02Multi-source signal agreement via Hellinger distance. Measures how much independent sources converge on the same prediction.
Calibration Scoring
$0.01Brier score and log score for evaluating prediction calibration. Measures how well predicted probabilities match actual outcomes.
Plan
Pathfinding and critical path analysis for sequencing tasks and finding optimal routes.
A* Pathfinding
$0.03A* shortest path with K-shortest alternatives using Yen's algorithm. Supports cost, time, and risk heuristics with weighted multi-objective search.