Anton – Python, LangChain, LangGraph

Integration and intelligence platform normalizing heterogeneous data from multiple hospitality system sources — PMS, POS, scheduling, and HR tools — into a unified data model. AI agent layer analyzes per-property operational data and generates structured optimization recommendations in real time. Multi-tenant architecture serving property managers, regional directors, and platform administrators.

• Built custom data integration architecture from scratch: ingestion pipelines normalizing heterogeneous hospitality system inputs into a unified PostgreSQL-backed data model using SQLAlchemy and Alembic migrations;
• Designed and implemented AI agent pipeline using LangChain, LangGraph, and OpenAI API: stateful agent workflows with Pinecone vector retrieval, source-specific chunking strategy, function calling, and structured output generation;
• Built async FastAPI backend with Pydantic-validated schemas for concurrent multi-property data processing;
• Implemented Auth0 authentication and RBAC with separate access layers for three user tiers;
• Built Streamlit-based management dashboard with configurable per-property KPI views and revenue risk signals.

All-in-one booking platform for motorhomes, RVs, boats, and houseboats covering the full rental lifecycle — asset discovery, identity verification, payment processing, booking management, and interaction history. Greenfield build from planning through production delivery under a fixed MVP deadline.

• Designed and built async FastAPI backend from scratch: SQLAlchemy data models, Pydantic-validated API layer, per-asset availability calendars, pricing rules, and booking state management;
• Built booking workflow engine: reservation creation, date conflict detection, availability locking, status transitions, cancellation, and refund logic;
• Implemented Celery + Redis periodic task infrastructure: booking reminders, availability expiry checks, payment capture triggers;
• Implemented secure payment integration with card-on-file, partial payment, deposit handling, and GDPR-compliant identity verification;
• Managed AWS infrastructure with Terraform (EC2, RDS, S3); wrote Pytest coverage across booking engine and payment flows.

Greenfield healthcare data management platform for regulated workflows — clinical cases, animal transfers, family trees, and medical report generation. Built from schema design through production deployment on AWS.

• Designed and built application architecture from scratch: SQLAlchemy data model, Alembic migration strategy, async API layer, RBAC, and audit trail for regulated data;
• Automated multiple manual workflows, significantly reducing operational overhead across facilities;
• Built report generation pipeline processing hundreds of thousands of records — general, medical, and financial reports per animal, group, and facility;
• Wrote Pytest test suite covering core data workflows; delivered complete technical documentation alongside production deployment.

Desktop application for design and simulation of photonic circuits with multi-user real-time collaboration and AWS serverless backend. Supports calculation-heavy design workflows, collaborative sessions, and cloud-based project sharing.

• Implemented Python business logic for calculation-heavy components: parameter adjustment, constraint validation, and geometry computation;
• Built real-time data synchronization layer across multi-user collaborative sessions with state persistence;
• AWS Lambda + DynamoDB + S3 backend for cloud file storage, project sharing, and role-based access;
• Wrote Pytest coverage for component CRUD, real-time sync, and computation logic.

Backend engineering across enterprise tools and B2B SaaS products — async API services, internal Streamlit-based dashboards and reporting tooling, Elasticsearch-powered search, and third-party API integrations across multiple product verticals.

• Designed and maintained async REST API services with FastAPI and SQLAlchemy across multiple product verticals;
• Built internal Streamlit tooling: data views, reporting dashboards, and workflow interfaces backed by FastAPI services;
• Implemented Elasticsearch and OpenSearch-based search and filtering across large datasets;
• Integrated third-party APIs and OAuth providers into application service layers;
• Maintained Docker + Nginx production environments with GitHub Actions CI/CD.