Seasoned Software Developer with a proven track record of engineering scalable backend infrastructures and intelligent, AI-driven applications. Blending precise system design with modern framework optimization, I specialize in building high-throughput REST/GraphQL APIs, handling complex database concurrency (ACID compliance), and implementing autonomous multi-agent RAG pipelines from the ground up. A firm believer in learning through experience.
- 💼 Current Role: Full Stack Developer at EY (Engineering robust solutions for enterprise scale).
- 🥅 Goal: To become a professional Software Developer.
- ⚡ Personal Goal: Bridging the gap between robust, enterprise-ready Java backend architectures and bleeding-edge local AI orchestration.
- Distributed Systems & Concurrency: Designing transactionally safe backend engines using Pessimistic Locking to eliminate concurrent write race conditions, and implementing the Saga Orchestrator Pattern to automate compensating transactions (refund loops) across distributed networks.
- Advanced API Architecture: Constructing robust, efficient API layers using traditional RESTful Controllers and modern GraphQL Engines—specializing in type-safe Schema Definitions, custom Query mapping, and non-blocking Resolver Logic to prevent over-fetching.
- Agentic AI & Multi-Agent Orchestration: Moving beyond basic RAG workflows by using Spring AI to coordinate autonomous Guardrail, Generator, and Evaluator agent loops that dynamically critique, score, and self-correct LLM responses in real-time.
- Vector Memory & Data Ingestion: Designing multi-tenant ingestion pipelines utilizing Apache Tika for text parsing, TokenTextSplitters for chunking, and ChromaDB (Docker-hosted) for indexing high-dimensional semantic vector embeddings (
nomic-embed-text).
🔍 View Detailed Architectural Breakdown
- Pessimistic Locking: Enforcing rigid database-level exclusive locks to preserve ACID compliance and prevent data anomalies during heavy concurrent updates.
- Saga Orchestration: Coordinating distributed state machines with structured error handlers to roll back data changes gracefully across isolated microservices when a network timeout occurs.
- GraphQL Resolvers: Writing structured data-fetching logic to streamline backend performance and optimize payload distribution.
- Spring Web Paradigms: Engineering robust API endpoints utilizing proper framework structures (e.g., handling complex multi-format payloads via
@RequestPart).
- Anti-Hallucination Loops: Building stateful evaluation cycles with bounded maximum-retry guardrails that programmatically audit model outputs against raw ground-truth text blocks before client delivery.
- Decoupled Data Ingestion: Automating localized text-parsing with fixed-size overlapping text windows to preserve deep semantic boundaries during vector transformation.
Random dev joke for you!
