Sai Shekhar Moola

Full-stack engineer with 5+ years of production experience across backend infrastructure and AI systems. Currently architecting AI-powered enterprise solutions at ServiceNow, building provider-agnostic LLM orchestration, multi-agent workflows, custom evaluation harnesses, and MCP integration layers that enable Fortune 500 companies to automate complex business processes at scale. Deep technical foundation in Go, Python, and TypeScript spanning microservices, databases, and cloud infrastructure.

Proven ability to own enterprise deployments end-to-end: from integration scoping and agent architecture through executive review, translating AI capabilities into measurable business outcomes.

Software Engineer (Backend and AI) - ServiceNow - Detroit, MI

(2024-07)

• Built AI-powered ticket classification system that reduced resolution time by 52%, generating $8.4M in annual operational savings; implemented provider-agnostic LLM routing (Anthropic Claude and OpenAI), category-specific agent workflows with approval gates, and Harness-driven zero-downtime model updates.
• Built custom evaluation harness from scratch tracking hallucination rate, output regression across model versions, and latency drift; reduced mean time to diagnose production issues by 34% and enabled non-technical customers to understand agent decisions through readable audit trails.
• Architected end-to-end multi-agent integration layer enabling enterprise customers to connect ServiceNow agents to third-party APIs and internal systems via MCP servers; owned deployments from integration scoping through executive review, cutting onboarding time from weeks to days.
• Designed and shipped A2A (agent-to-agent) communication protocol enabling specialized sub-agents to delegate tasks, share context, and resolve conflicts autonomously, reducing manual orchestration overhead for complex multi-step enterprise workflows across 8,000+ enterprise customers.
• Engineered FastAPI-based MCP server infrastructure with native asyncio support, enabling agent-to-API calls at 175ms p95 latency while maintaining strict SLA compliance across ServiceNow's multi-tenant platform serving 8,000+ enterprises.
• Built knowledge-grounded RAG pipelines using pgvector and PostgreSQL, improving Now Assist retrieval accuracy by 22% and reducing inference latency from 420ms to 190ms; communicated results directly to enterprise customers in executive reviews.
• Partnered with enterprise customer success and product teams to translate agent reliability data and model performance trade-offs into roadmap priorities, owning the feedback loop from production deployment to product iteration.
• Delivered AWS integration connectors (Lambda, S3, EC2) enabling customers to automate cloud infrastructure workflows directly from ServiceNow, reducing manual cloud administration overhead for enterprise engineering teams.
• Containerized AI microservices and agent inference workloads using Docker and Kubernetes, supporting multi-instance deployment across 8,000+ enterprise customers globally.

Software Developer - Orion TechnoLab - India

(2019-09 - 2022-11)

• Hardened a Spring Boot order status service through a holiday sales peak, optimizing thread pool and connection pooling to sustain 800K+ daily requests below 0.1% error rate.
• Introduced gRPC for inter-service communication across order, inventory, and pricing microservices, cutting serialization overhead by 38% at peak load.
• Built Elasticsearch-backed product search with fuzzy matching and faceted filtering, reducing average search response time from 2.1s to 310ms across a catalog of 2M+ SKUs.
• Designed a Kafka-based order event pipeline decoupling checkout, inventory reservation, and notification services, eliminating synchronous dependencies and reducing order processing latency by 29%.
• Rebuilt product listing and checkout UI in Angular with OnPush change detection and lazy-loaded route modules, improving Lighthouse score from 41 to 76 and cutting time-to-interactive from 6.2s to 2.8s on mobile.
• Secured the platform API layer using Azure API Management with JWT authentication and circuit breakers, cutting unauthorized access attempts by 94%.

M.S. - Computer Science - University of the Pacific (2023-01 - 2024-12)