Yaroslav R.

A multi-tenant SaaS platform for service and maintenance companies, centralizing customer management, service requests, scheduling, estimates, invoicing, and payments, with AI-assisted request processing and estimate drafting. Led technical discovery with client stakeholders, translating requirements into a scalable SaaS architecture. Designed multi-tenant modules for organizations, customers, and technician assignments using TypeScript, NestJS, and PostgreSQL. Built the Next.js/React operations dashboard for request and job tracking. Implemented REST API contracts and secure RBAC workflows for owners, dispatchers, and technicians. Architected real-time updates via WebSockets, Redis, and BullMQ. Integrated OpenAI API and Stripe for AI-assisted processing and payments. Supported AWS deployments with Docker, and mentored engineers on architecture.

A developer-focused SaaS solution consisting of a web analytics dashboard and a cross-platform mobile application for monitoring production AI systems, processing high-cardinality telemetry data with near real-time visibility into agent runs, nested traces, prompts, tool calls, token usage, latency, costs, errors, and evaluation results. Contributed to backend architecture decisions and developed a high-throughput telemetry ingestion service using TypeScript, NestJS, and Fastify. Designed scalable ingestion workflows with batching, idempotency, and rate limiting through Redis and BullMQ. Used PostgreSQL for transactional metadata and ClickHouse for high-volume storage and analysis of traces and costs. Built aggregation and retention workflows for fast analytical queries. Implemented nested traces, prompt versioning, automated evaluations, and PII masking. Developed a React Native incident-monitoring app and participated in architecture discussions, code reviews, and AWS deployments.