NBA • Pre-Match .NET 9 • Razor Pages Fair Lines Engine Agentic AI

BettingOdds

An enterprise-grade NBA pricing platform that generates fair moneyline, spread, and total lines using advanced stats + player relevance — with an internal coding agent system for rapid development.
Open Sync Center Upcoming Games Agents Dashboard
Moneyline Spread Total (O/U) Player-adjusted Audit-ready
Architecture
Clean layered
UI (Pages) → App → Domain → Infra/Data. Pricing math lives only in App/Pricing.
Core product
Fair lines engine
Baseline projection + player-adjusted factors + confidence scoring.
Agentic AI
Planner→Impl→Review
DB-backed runs, patch sets, guard reports, build & test logs.
System console
Production-ready posture
SQL Server
Data source stats.nba.com
Storage EF Core Migrations
Execution Build + Test runner
Pricing contract Protected
Open Agents (runs + patches)
Review persisted runs, apply patch sets, and run build/tests.
Positioning
Find edges faster
Produce a fair line, compare to the market, and identify mispriced spots — with transparent model inputs.
Audit Repeatable Traceable
Data
Run All
Sync Center
Refresh teams, season context, advanced snapshots, and schedule in a controlled pipeline.
Recommended start.
Pricing
No IDs
Upcoming Games
Browse the next 3 days, generate fair lines, and open matchup details in one click.
Best operational view.
Platform
AAI
Agents
Run Planner→Implementer→Reviewer, inspect patch sets, and manage changes with audit trails.
Enterprise automation layer.
League
Catalog
Teams
Explore teams and underlying metadata that supports projections and roster logic.
Audit
Snapshots
Team Stats
Review the latest advanced stats snapshot currently feeding the pricing model.
Operational workflow
How this platform is used
1) Sync & snapshot
Run Sync Center to refresh teams, stats snapshots, and the schedule for upcoming games.
2) Price the slate
Use Upcoming Games to project spreads/totals/moneylines and open detailed matchup pricing.
3) Iterate safely
Use Agents to generate guarded patches, apply changes with audit trails, then build + test.
This is real engineering: clean architecture, protected pricing contracts, DB-backed agent runs, and reproducible pipelines.