# TAXIE Business Strategy Framework | Section | Purpose | Key Components | Core Functions | Implementation Details | | ------------------------------------- | ------------------------------------------------------- | ------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | 1. 🗺️ Business Domain Representation | Models TAXIE's business environment and market position | BusinessPose, BusinessControl, MarketSpace, TaxieStartupEnvironment | distance_to_market_segment(), physical_step(), integrate_business_path() | BusinessPose tracks position (x=range, y=earnings) and heading (strategic direction); MarketSpace defines market boundaries with (150-300 miles range, 0−3000earningsadvantage)andcompetitorregions(Hertzfleetat220−260miles,0-3000 earnings advantage) and competitor regions (Hertz fleet at 220-260 miles, 0−3000earningsadvantage)andcompetitorregions(Hertzfleetat220−260miles,1400-1800); Environment integrates the physics of business movement with constraints | | 2. 🧠 Probabilistic Models | Provides generative models for business strategy | Sensor models, step models, path models, prior distributions | sensor_model_one(), step_model(), full_model(), uniform_pose_prior(), localized_prior() | Sensor model uses genjax.normal with market distance + noise; Step model implements mv_normal_diag for uncertain movement; Full model combines steps with sensors; Priors include uniform (whole market), mixture (market segments), and localized (around specific position) | | 3. 🔄 Inference Methods | Implements algorithms for reasoning under uncertainty | Noisy sensing, importance resampling, SIS with rejuvenation, grid-based rejuvenation | noisy_sensor(), importance_resample(), SISwithRejuvenation class, grid_fwd_proposal(), grid_bwd_proposal() | SIS scan function tracks particles through steps; Grid rejuvenation uses SMCP3 with forwards/backwards proposals; Make grid functions create exploration spaces; Grid margin set to 0.5 and resolution to 5 | | 4. 🕹️ Visualization and Dashboard | Creates interactive exploration tools | Market visualization, path plots, interactive widgets, dashboard | plot_market_space(), plot_path_with_confidence(), pose_widget(), camera_widget(), taxie_dashboard() | Market visualization uses colored regions for competitors; Path plots show business trajectory with confidence intervals; Dashboard combines Market Sensing tab (for exploration) and Business Path tab (for analysis); Pose widgets enable dragging to explore market positions | | 5. 📊 Business Insights | Extracts actionable recommendations | Analysis functions, strategy interpretation, metrics calculation, insight generation | analyze_taxie_business_strategy(), interpret_business_strategy(), taxie_business_insights() | Analyzes two scenarios: confident (low noise) vs uncertain (high noise); Calculates key metrics: error from optimal position, position variance, probability of profitability (reaching 260+ miles range with $1500+ earnings), competitor collision risk; Final recommendation to shut down operations based on high maintenance costs and Hertz's competitive advantage | [gist code](https://gist.github.com/hyunjimoon/daaddeb97629ee54a400996d584df773)