
MediBot Agentic RAG
Agentic AIAgentic RAG medical assistant with tool routing, memory, retrieval, safety controls, and adversarial evaluation.
I build secure, reliable systems with Java, Python, Spring Boot, FastAPI, Docker, AWS, and applied AI patterns, from financial services integrations to agentic RAG systems with evaluation, memory, and safety controls.
1M+
daily financial transactions supported
Java and Spring Boot services for fintech and core-banking integrations.
99.9%
service uptime target
Reliability work across backend services, caching, and deployment workflows.
50%
manual workflow reduction
Automation across backend, data, and QA workflows.
60
adversarial AI evaluation cases
MediBot safety, retrieval, routing, and out-of-scope verification.
Four current projects that show backend reliability, secure identity, applied AI orchestration, evaluation, and production readiness.
Experience and impact
Software Engineer
Built Java and Spring Boot services supporting high-volume financial transaction workflows, payment integrations, authentication, caching, and CI/CD.
Software / QA Automation Engineer
Improved regression coverage and deployment reliability through automation, backend workflow validation, and cross-functional release support.
Graduate Teaching Assistant
Taught and mentored students in programming and systems concepts while building grading and instructional support workflows.
Lead Software Engineer
Lead software engineer for commerce workflows, product implementation, technical planning, and delivery quality.
Engineering capabilities
Resource-oriented APIs, service boundaries, validation, error handling, caching, and production-minded configuration.
OAuth 2.0, OIDC/PKCE, Keycloak, JWT validation, authorization boundaries, and organization-scoped data.
LangGraph agents, tool routing, retrieval, memory, safety fallbacks, and deterministic local verification.
PostgreSQL, Oracle, SQLite, schema migrations, SQLAlchemy, Hibernate, and testable persistence layers.
Docker, Docker Compose, GitHub Actions, integration testing, health checks, and reproducible local startup.
Prompt-injection defense, classification metrics, failure analysis, false-negative review, and safety reporting.
About
My strongest work combines backend reliability, secure identity, data modeling, applied AI orchestration, and evaluation. I like systems where the architecture can be inspected and the results can be reproduced.
Contact