Francis Beeson

llm_benchmarks. A collection of benchmarks and datasets for evaluating LLM.

ai-dev-workflow. Structured workflow prompts for AI-assisted software development. Guides engineers through PRD creation, architecture design, task management, and detailed AI development tool specifications for robust code generation.

single-file-ai-agent-tutorial. Python version of "How to Build an Agent" by Thorsten Ball

llm-driven-deterministic-modelling. Insights into how to use LLM for Program Control Flow and Steerable Agents.

pen-test-llm. A service for using uncensored LLMs for penetration testing of T&S models.

rag-insights. A collection of ideas, insights, and resources for improving RAG implementations.

query-preprocessing-workflow. Preprocesses queries to check if it can be processed.

ai-assistants-guide. In-depth exploration of various AI assistant technologies, detailing their definitions, key features, capabilities, and use cases. It includes a comparative analysis of different AI systems, such as semantic search, retrieval-augmented generation (RAG), conversational agents, copilots, and agentic virtual assistants.

html-ingestor. A Python package for ingesting HTML documents while preserving semantic structure to generate RAG-friendly input documents and chunks.

simple_language_detection. Calls cld3 to detect languages of all pdf/text files in a folder.

key-mwe. Extracts keywords and multi-word expressions (MWE) using within corpus PMI, and estimates keyword and MWE Keyness using between corpora PMI.

second-brain-synapse. LLM-driven backend for organising and powering GTD-like Second Brain systems.

cs_chat_data_insights_generator.

linkedIn-job-tracker. Bot for analysing and tracking user-relevant job postings.

openFDA-labels. Anlysis of Ingredients on Medication Labels