yash176a/pricing-engine

How yash176a/pricing-engine works

This is a synthetic-market dynamic pricing engine: it generates noisy historical observations, fits a demand model, and selects prices to maximize modeled profit. Its evaluation scores the optimized policy against the market’s true demand function, then can register and promote the trained model. The serving API resolves the production model for demand predictions and price recommendations. The graph follows the sampled training, evaluation, registry, optimizer, and service code; other README capabilities are included only where documented, and no unsampled implementation details are assumed.

Main entry points: train.py Training and evaluation · service.py FastAPI service · optimizer.py Price optimizer · run_eval.py Evaluation runner

Components 12

Market and evaluation

Synthetic market src/pricing_engine/data.py
Training and evaluation src/pricing_engine/train.py
Evaluation runner eval/run_eval.py
True demand function src/pricing_engine/data.py

Pricing models

Demand model src/pricing_engine/demand_model.py
Price optimizer src/pricing_engine/optimizer.py
Runtime settings src/pricing_engine/config.py

Model registry

Versioned model registry src/pricing_engine/registry.py
Registry files src/pricing_engine/registry.py

Serving interface

Command-line interface src/pricing_engine/cli.py
FastAPI service src/pricing_engine/service.py

Other components

Operator / API client

Connections 18

Operator / API client → Command-line interface: runs commands README.md
Operator / API client → FastAPI service: sends requests README.md
Command-line interface → Evaluation runner: runs evaluation src/pricing_engine/cli.py
Command-line interface → FastAPI service: starts server src/pricing_engine/cli.py
Evaluation runner → Training and evaluation: trains and evaluates eval/run_eval.py
Evaluation runner → Versioned model registry: registers and promotes eval/run_eval.py
Training and evaluation → Synthetic market: generates market src/pricing_engine/train.py
Training and evaluation → Demand model: fits model src/pricing_engine/train.py
Training and evaluation → Price optimizer: chooses prices src/pricing_engine/train.py
Training and evaluation → True demand function: scores policies src/pricing_engine/train.py
Synthetic market → True demand function: generates observations src/pricing_engine/data.py
Price optimizer → Demand model: queries predictions src/pricing_engine/optimizer.py
Versioned model registry → Registry files: persists artifacts src/pricing_engine/registry.py
FastAPI service → Versioned model registry: resolves and loads src/pricing_engine/service.py
FastAPI service → Price optimizer: optimizes prices src/pricing_engine/service.py
FastAPI service → Demand model: predicts demand src/pricing_engine/service.py
Runtime settings → Training and evaluation: sets training parameters src/pricing_engine/train.py
Runtime settings → Price optimizer: sets price bounds src/pricing_engine/optimizer.py

Drawn by GitDiagram on November 12, 2026 from the repository's file tree, README and source files. Read as Markdown ·