Key Skills

Systems & Architecture

• Designed and deployed a multi-service platform using Proxmox VE and LXC containers
• Service separation: abstraction layer, backend, frontend, cache, internal DB, reverse proxy
• Container boot order planning and service dependency handling
• Internal HTTP routing and reverse proxy setup for clean access patterns
• Basic operational setup: service startup scripts, clean restarts, and isolation between components

Backend Development & APIs

• Built REST endpoints using FastAPI for query routing and analysis execution
• Structured request/response handling for passing config, ranges, and analysis settings
• Input validation and defensive error handling to avoid bad requests breaking services
• Logging of key actions (requests, failures, cache events) for traceability

Data Engineering & Database Integration

• Dynamic database access via a custom abstraction layer (ODBC + REST patterns)
• Worked with SQL Server and PostgreSQL connection handling and query execution
• Designed a consistent data flow to return query results as pandas DataFrames / JSON payloads
• Internal PostgreSQL used for configuration storage and audit-style logging

Caching & Performance

• Implemented Redis caching for raw query reuse and repeat analysis speed-up
• Key-based retrieval for loading historical queries and processed outputs
• TTL handling and cache inspection for debugging and memory awareness
• Reduced repeated load on external databases by avoiding duplicate queries

Time Series Analysis

• Implemented Dynamic Time Warping (DTW) for comparing time ranges or batch signals
• Normalisation and scaling approaches used to support fair comparisons
• Traceback path logic through the DTW cost matrix for alignment interpretation
• Built the analysis pipeline to accept typed inputs and return structured results

Data Handling & Preprocessing

• Built a data dictionary approach to classify incoming columns (datetime, numeric, categorical)
• Datetime parsing and cleanup for mixed precision timestamps
• Missing data handling using safe defaults, warnings, and guard checks
• Standardised typed outputs so downstream analysis modules can assume stable inputs

Frontend Integration (Engineering-Focused UI)

• Dash frontend used to select databases, tables, time windows, and analysis methods
• Frontend-to-backend workflow design: query → cache → process → render → export
• CSV export support for processed results and repeatable analysis outcomes
• State handling for retrieving prior Redis keys and restoring previous work

Engineering Practices

• Modular Python structure: reusable classes, separated responsibilities, clean interfaces
• Practical debugging across distributed services (frontend, backend, Redis, DB)
• Worked with realistic dataset sizes (tens to hundreds of thousands of rows)
• Built with audit and compliance thinking in mind (logging, traceability, internal config DB)

What This Shows

This project shows end-to-end capability: from deploying services and wiring them together, through database access and caching, to time-series analysis and a working web UI that can be used to run repeatable investigations on industrial-style data.