The tools behind connected systems.

Claude models for code, analysis and structured reasoning, used interactively and through the Anthropic API.

Experience: AI-assisted development (including this website), workflow analysis, documentation and drafting.

Works with: Code editors, Docs, REST APIs

GPT models for drafting and research, and for assistants that call defined functions.

Experience: Drafting, research, prompt design and planning for marketing and product work.

Connecting language models to external APIs, structured business data and controlled tools to execute multi-step tasks.

Experience: Connected language models to app prototypes through Python and REST APIs, in local and hosted modes.

Works with: Python, REST APIs

Instructions and output schemas that make model responses predictable enough for other systems to validate and use.

Experience: Structured prompts for content, classification and assistant behaviour.

Evaluating and configuring locally hosted language models (Llama, Mistral and Qwen families) for application-specific inference, within the limits of the available hardware.

Experience: Ran Llama- and Mistral-family models locally through Ollama while prototyping an offline-capable assistant. Qwen-family models are being evaluated.

Works with: Python, Ollama API

Speech-to-text in voice assistant prototypes.

Voice output in assistant prototypes.

Exploring Grok-powered agents, xAI integrations and tool-connected workflows for research, analysis and operational automation.

Experience: Testing Grok for research and analysis, and the xAI API for tool-connected prototypes. A Grok subscription and xAI API access are separate: the API has its own keys and usage billing.

Works with: xAI API, REST APIs

Configuring Grok-based bots with defined instructions and connected workflows, separate from direct xAI API integration.

Experience: Configuring bot behaviour and testing how it fits into research and content workflows.

Designing coordinated agent workflows that separate research, execution, validation and human approval responsibilities.

Experience: Role-separated agent workflows designed and documented as blueprints in Build Your Stack. Not yet running for a client.

Automation architectures that combine autonomous task execution with validation, exception handling and human oversight.

Experience: Designing agent workflows with human approval steps, documented as the blueprints in Build Your Stack.

Agents that start from a webhook, a new record or a schedule instead of waiting for someone to open a chat.

Experience: Testing scheduled and event-driven runs for briefing and follow-up workflows.

Deciding what an assistant keeps between sessions, what it retrieves on demand and what it should never store.

Experience: Explored how assistants keep useful context between sessions.

Comparing models on task quality, latency, context length, licence terms and running cost before choosing one for a workflow.

Experience: Comparing hosted and open-weight models on the same prompts during prototyping.

Connecting controlled language-model inference with document retrieval, embeddings and permission-aware knowledge access.

Experience: Designed as the Internal Knowledge Assistant blueprint: answers drawn only from approved documents, with access following existing permissions.