Senior Software Engineer | Python & AI Engineer · LLM Systems · RAG · Search & Retrieval · Cloud

Building production-grade AI systems with Python — from intelligent APIs and agentic workflows to grounded retrieval, document intelligence, and scalable cloud infrastructure.

I'm a senior software engineer focused on Python, AI/LLM systems, RAG, semantic search, vector retrieval, document intelligence, and scalable backend architecture.

My recent work has centered on building production AI systems that do more than call a model — they retrieve the right evidence, validate structured outputs, orchestrate multi-step workflows, and return grounded answers with traceable citations.

• Building production APIs with Python, FastAPI, Pydantic, SQLAlchemy, and AsyncIO, with additional backend experience in Go, C#, .NET, and Node.js
• Designing RAG pipelines, agentic workflows, semantic search, and vector retrieval with frameworks like Pinecone, Qdrant, and Milvus
• Integrating Anthropic Claude, OpenAI, and Google Gemini
• Building document extraction, classification, enrichment, and search workflows
• Designing high-throughput backend services and distributed systems
• Working with PostgreSQL, MongoDB Atlas, vector databases, and retrieval ranking
• Deploying containerized AI services with Docker, Kubernetes, AWS, Terraform, and cloud infrastructure
• Supporting AI products with React, Next.js, and TypeScript when full-stack delivery is needed

┌──────────────────┐
│ Documents / Data │
└────────┬─────────┘
         │
         ▼
┌──────────────────┐
│ Parsing /        │
│ Enrichment       │
└────────┬─────────┘
         │
         ▼
┌──────────────────┐
│ Embeddings +     │
│ Search / Ranking │
└────────┬─────────┘
         │
         ▼
┌──────────────────┐
│ Retrieval / RAG  │
└────────┬─────────┘
         │
         ▼
┌──────────────────┐
│ LLM / Agentic AI │
└────────┬─────────┘
         │
         ▼
┌──────────────────┐
│ Grounded Answer  │
│ + Citations      │
└──────────────────┘

• AI-powered search and document intelligence platforms
• RAG systems with semantic search and vector databases (Pinecone, Qdrant, Milvus, pgvector)
• Agentic workflows with multi-step search and tool use
• Citation-grounded LLM applications
• Structured extraction and classification pipelines
• High-throughput Python and Go APIs and asynchronous services
• Data normalization, ranking, deduplication, and enrichment workflows
• Production AI systems deployed on AWS with Terraform and containers

Built and enhanced an AI-powered regulatory intelligence platform using Python, FastAPI, Pydantic, SQLAlchemy, AsyncIO, Anthropic Claude, semantic search, and vector search.

Key contributions:

• Built multi-step agentic retrieval workflows
• Integrated structured LLM outputs for document and reference extraction
• Designed related-document discovery and enrichment pipelines
• Implemented citation-grounded answers to reduce unsupported responses
• Improved search quality with fuzzy matching, recency scoring, and metadata boosts
• Separated LLM enrichment from deterministic search services
• Added async processing, validation, error handling, and automated tests

Worked on high-volume gaming and transactional platforms including Crash, TurboGames, Slots, Jackpot, and WinBoxes.

Key contributions:

• Optimized high-volume bet ingestion and stored procedures
• Improved platform scalability for transaction-heavy workloads
• Supported modular FastGame architecture and independent deployments
• Modernized legacy components and upgraded SDK integrations
• Improved maintainability, release flexibility, and production reliability

Built backend infrastructure for real-time trading and brokerage platforms.

Key contributions:

• Designed high-throughput APIs and transaction workflows
• Integrated external financial and liquidity-provider APIs
• Worked on low-latency and concurrent processing systems
• Optimized database access and real-time data flows
• Built reliable distributed backend services

Developed Python automation for geospatial and surveying workflows using Python, ArcPy, and ArcGIS SDK.

Key contributions:

• Automated repetitive spatial-data processing
• Built reusable data-transformation and validation utilities
• Developed batch-processing workflows for geospatial datasets
• Improved engineering productivity through Python tooling

I focus on AI systems that are:

• Grounded — responses should be supported by retrieved evidence
• Reliable — structured validation, testing, error handling, and fallbacks matter
• Scalable — retrieval and model workflows should handle growing data and traffic
• Observable — production AI needs logs, metrics, and traceable behavior
• Maintainable — probabilistic AI logic should be separated from deterministic application logic
• Useful — AI should solve real workflow problems, not just demonstrate model capabilities

I'm open to collaborating on Python and Go backend systems, AI/LLM applications, RAG, semantic search, vector databases, document processing, retrieval architecture, and production AI infrastructure on AWS.

Building reliable AI systems with Python, retrieval, and grounded intelligence.