Rayan
Chahid

Built production systems for payments reliability and autonomous CI/CD remediation, combining deterministic orchestration with AI-assisted investigation and verification.

• • Owned a zero-downtime runtime-configuration migration for a Tier-1 payments service handling millions of daily transactions, reducing change lead time from 31 hours to under 5 minutes through a governed configuration store, feature flags, and automatic fallback.
• • Architected an autonomous CI/CD remediation platform that detects failing test suites, investigates root causes with parallel AI agents, authors fixes, independently verifies changes, and creates review-ready code changes with human approval before merge, saving 10+ engineer-hours per week.
• • Built a deterministic orchestration runtime with typed contracts, process isolation, bounded retries, durable state, keyed synchronization, and resource-aware concurrent scheduling, reducing manual test-remediation time by 90%.
• • Restored multiple integration-test suites from a 100% YTD failure rate to passing through a multi-stage verification gate with selective test reruns, regression detection, test stability analysis, and informed retries.

Building Sentinel, a policy-constrained AI agent for conversational financial analysis and fraud investigations.

• • Built a policy-constrained AI agent using intent routing, tool permissions, and finite-state orchestration for conversational financial analysis and fraud investigations.
• • Integrated live transactions, Gmail, Calendar, Telegram, Tavily web search, and opt-in on-device telemetry to ground fraud investigations and answer natural-language questions about spending, schedules, and transactions.
• • Engineered deterministic financial analytics and evidence retrieval services with idempotent case transitions and replay protection for reliable multi-turn investigations.

Led two engineering teams from product definition through deployment, translating project ideas into scoped roadmaps and coordinating sprints to ship two full-stack applications.

• • Led two engineering teams (10 developers total) from product definition through deployment, translating project ideas into scoped roadmaps, assigning ownership, conducting code reviews, and coordinating sprints to ship two full-stack applications.
• • Directed technical development of a prerequisite-graph course planning engine and a course/professor analytics platform, spanning catalog ingestion and normalization, graph-based trajectory planning, data pipelines, and APIs.