Ridha Tech
Get Started
MVP: 97% Complete | Launch: Q2 2026

SIPAP: Sports Intelligence Platform

AI-powered probability assessment delivered via WhatsApp. Five specialized agents analyze statistical models, machine learning predictions, team form, market sentiment, and news intelligence to identify positive expected value opportunities.

The Problem We Solve

Betting markets are complex, data is scattered across multiple sources, and existing platforms operate as black boxes. Users need transparent intelligence, not just odds.

Scattered Data Sources

Sports data is fragmented across dozens of platforms. Aggregating insights manually is time-consuming and error-prone.

Black Box Predictions

Existing platforms provide predictions without explanation. Users have no way to validate the reasoning or understand confidence levels.

High Friction Access

Mobile apps require downloads, registrations, and permissions. Users in emerging markets need zero-friction access.

No Expected Value Analysis

Most platforms show odds without comparing them to true probabilities. Users can't identify where the market is mispriced.

Our Solution

SIPAP combines five specialized AI agents with WhatsApp-first delivery to provide transparent, evidence-based sports intelligence.

Multi-Agent Intelligence

Five specialized AI agents work together: Statistical Model Agent, ML Prediction Agent, Team Form Agent, Market Sentiment Agent, and News Intelligence Agent.

Ensemble predictions with confidence scoring

WhatsApp-First Interface

Zero app downloads required. 80-95% WhatsApp adoption in target markets means instant access for millions of users.

Natural language queries via conversational AI

Expected Value Analysis

Compare our probabilities vs market odds to identify +EV opportunities. Calculate edge, assess risk, and show confidence levels on every prediction.

Only recommend bets with positive expected value

Transparent Reasoning

Every prediction shows evidence sources, agent-by-agent breakdown, quality gates, and ensemble confidence. Users see exactly how we arrive at conclusions.

No black boxes - complete explainability

Technical Architecture

Production-grade AWS infrastructure orchestrating 60+ resources for sub-100ms response times.

User → Twilio WhatsApp → API Gateway → SQS FIFO
ECS Fargate (daemon) → Multi-agent orchestrator
5 AI Agents (Statistical, ML, Form, Market, News)
Data Layer: 3 free-tier APIs + Aurora PostgreSQL + Redis
Intelligence Layer: Weather, news sentiment (Claude/Bedrock)
60+
AWS resources
<100ms
Latency
72/72
Tests passing
0
Quality errors

Development Journey

Phase 0-2: Foundation

Common Libraries & Data Infrastructure

Built sipap-common (shared utilities), sipap-serverlesshandler-mcp (MCP base classes), and sipap-batch-scraper (API-based data collection). Achieved zero quality gate errors: 0 mypy errors, 0 ruff errors, 85%+ test coverage.

Phase 3-4: MCP Servers & Intelligence

Data Layer & AI Integration

Deployed 5 MCP servers (Sports Data, Odds Intelligence, News Intelligence, Weather Data, Prediction Engine). Integrated Claude AI via AWS Bedrock for natural language processing and sentiment analysis.

Phase 5-6A: Orchestrator & WhatsApp

Multi-Agent System & Interface

Built sipap-master orchestrator coordinating 5 specialized agents. Integrated Twilio WhatsApp Business API for conversational interface. MVP technically complete at 97% - waiting for business verification to go live.

Ready to Launch

SIPAP MVP is 97% complete. Launching Q2 2026 with global soccer coverage.

View Technology StackRead Build Journal