Software engineerbuilding enterprise apps,tools and AI.
I design and build enterprise software end to end: full-stack apps, backend systems, integrations, and the AI tooling around them.
Apps,tools,andAI.Builtendtoend.Applauseoptional.
From chatbots to platforms.
Shipped custom software for clients.
Built and tested solutions shaped around each client's requirements.
Taught chatbots to answer.
Dialogflow responses, third-party API integrations and agent routing, tested end to end.
Modernized a migration platform for 50+ environments.
A legacy Oracle APEX and Java tool became a React, TypeScript and Flask platform.
Moved millions of records, retryably.
An Oracle data model with chunked, independently retryable batches: hundreds of thousands of records typically, millions at peak.
Made 5,000+ assets searchable by meaning.
Semantic and vector search with RAG lifted overall search performance by 35%.
Put a Claude agent into production.
Intent classification, session memory, tool calling and guardrails, with cost-aware model routing.
Built the platform six teams share.
Project Almanac guides AI-assisted development, with productivity gains of up to 30%.
Selected work.
- Insight Agent5,000+
- Project Almanac6 teams
- Intelligent Data Migration50+
- AI Workflow Automation
- Flowchart Generator
- Python
- Java
- TypeScript
- React
- Spring Boot
- Flask
- Oracle Fusion ERP
- Oracle Cloud Infrastructure
- REST APIs
- SOAP APIs
- JWT
- Playwright
- Docker
- Anthropic Claude API
- RAG
What itadded upto.
What it added up to
5,000+
Indexed for semantic search (CDP)
35%
Improvement in search performance
6 teams
Using Project Almanac for AI-assisted development
Up to 30%
Productivity improvement with AI workflows
50+
Oracle environments supported (IDM)
How I work.
Own it end to end.
On Project Almanac: architecture, backend, frontend, AI orchestration, deployment and testing.
Leave it testable.
Playwright, JUnit, pytest and Postman across unit, integration, API, regression, load and security tests.
Secure it from the start.
JWT replacing HTTP Basic, input validation, API authorization and vault-managed secrets, plus prompt-injection detection on the AI side.
Measure what it costs.
A token-usage dashboard tracked cost, errors and quotas, and per-tier routing sends basic requests to lower-cost models.
Read the whole story.
Software Engineer with experience designing and delivering enterprise full-stack, backend, and AI-enabled applications.