James Murray is a programmer, software designer, web developer and AI knowledge-systems builder based in Kelowna, British Columbia, Canada. With more than 40 years of hands-on experience, he has worked across the full evolution of modern computing—from early custom software and database systems to large-scale websites, Python automation, secure APIs, vector databases and production AI search.

His strength is not limited to one technology or one stage of a project. James works across the entire system: defining the problem, designing the architecture, collecting and cleaning data, writing the processing tools, structuring the knowledge, building the vector database, creating the API, integrating the public website, securing the service, testing the results and documenting how to recover the system if something fails.

From Early Software to Production AI Systems

James began programming during the early personal-computer era and has continued adapting through every major shift in technology. His career has included custom software, databases, reporting systems, website architecture, PHP applications, content platforms, hosting systems, automation, search, SEO, media production and now private AI knowledge platforms.

• Programming and software design: more than four decades of practical development, troubleshooting and product building.
• Web development: PHP, HTML, CSS, JavaScript, MySQL, MariaDB, hosting architecture and large content-driven websites.
• Data engineering: Python and PowerShell pipelines for crawling, extraction, cleanup, normalization, deduplication and structured exports.
• AI and vector search: embeddings, retrieval-augmented generation, Weaviate collections, semantic search, metadata filtering and grounded answers.
• Infrastructure: Linux VPS administration, SSH, FastAPI, Uvicorn, systemd, health checks, logging, backups and rollback environments.

AddictionTube and the Future of Private AI Search

James is the creator and technical architect behind AddictionTube, a large recovery-information platform that has evolved from a traditional website network into a private AI knowledge system.

The AddictionTube therapist-search platform demonstrates his ability to take a complex real-world project into production. The system includes:

• 6,506 therapist profiles organized into 26,725 searchable knowledge chunks.
• Coverage across 474 Canadian cities and communities.
• Python systems for website discovery, crawling, content extraction, data cleanup and JSONL generation.
• A Weaviate vector database with profile-specific metadata and structured knowledge chunks.
• FastAPI services for semantic search, city browsing and therapist-specific question answering.
• PHP proxy integration on shared hosting using cURL, request validation and protected environment files.
• Shared-secret authentication, HMAC visitor tokens, AES-GCM image tokens and per-visitor rate limiting.
• Structured logging, health checks, staging services, rollback services, backups and reconstruction documentation.
• A completed 100/100 API regression test covering supported, unsupported, crisis and location-sensitive questions.

This work represents more than a chatbot or search box. It is a complete production architecture that transforms raw website information into a private, controlled and source-grounded knowledge system.

Recovery Experience and Industry Understanding

James also brings more than 18 years of sobriety and over two decades of experience connected to addiction treatment, recovery information and provider marketing. This lived and professional experience gives him a practical understanding of the questions families ask, the information treatment providers often fail to explain and the importance of clarity, safety and trust in healthcare-adjacent systems.

That combination of recovery experience and technical depth allows him to build systems that are not only functional, but also grounded in the needs of real people making difficult decisions.

Technical Capabilities

• Languages and tools: Python, PHP, PowerShell, JavaScript, HTML, CSS and SQL.
• Data formats: CSV, JSON, JSONL, SQL imports and structured lookup files.
• Databases: Weaviate, MySQL, MariaDB and custom structured datasets.
• API and AI stack: FastAPI, Uvicorn, REST APIs, OpenAI models, embeddings and RAG workflows.
• Server operations: Linux, SSH, remote administration, systemd, ports, environment files, journal logs and service recovery.
• Production security: authentication headers, HMAC, AES-GCM, protected images, rate limiting and privacy-safe logging.
• Quality assurance: automated regression testing, pagination tests, duplicate detection, ordering validation, backups and checksums.

94-Project Technical Portfolio

James has created a portfolio of 94 project pages covering AI knowledge systems, vector databases, recovery technology, software engineering, business innovation, digital strategy, identity systems and technical leadership.

The complete executive portfolio is available here:

View James Murray's AI Knowledge Systems and Technical Portfolio

Consulting, Contract and Technical Leadership

James is available for consulting, contract development, technical architecture, data engineering, AI knowledge-system implementation, vector-database development, secure API integration and production rescue work.

He is particularly well suited to projects that cross several technical boundaries and require someone who can understand the full problem, build the system and get it running.

Contact James Murray