Pinned Loading

1. internal-ai-agent-eval-lab internal-ai-agent-eval-lab Public

   Public benchmark and evaluation harness for AI-agent safety and reliability across RAG grounding, safe refusal, prompt-injection resistance, tool governance, observability, and CI.

   Python 2

2. llm-redteam-harness llm-redteam-harness Public

   LLM red-team evaluation harness for testing prompt-injection, refusal, leakage and safety behaviours with structured prompts, scoring and reproducible reports.

   Python

3. event-extraction-llm-baseline event-extraction-llm-baseline Public

   Zero-shot LLM event extraction baseline using Qwen2.5-7B-Instruct on MAVEN and WikiEvents, with constrained prompting, JSON evaluation and A100 HPC inference.

   Python

4. neobank-product-analytics neobank-product-analytics Public

   Product-ready synthetic fintech growth and pricing intelligence platform using dbt, DuckDB, BigQuery, Cloud Run, Streamlit, FastAPI, experimentation, activation modelling, geo-lift, pricing analyti…

   Python

5. uk-property-analytics uk-property-analytics Public

   Analytics engineering portfolio project using dbt-core, DuckDB and Streamlit on 4.99M UK Land Registry records, with 88 data tests, CI, docs and dashboard.

   Python

6. marketing-effectiveness-lab marketing-effectiveness-lab Public

   Marketing effectiveness analytics lab for MMM, causal measurement, budget optimisation, and commercial decision support.

   Python