.png){kind=link}
Challenge 4
The team built an automated Lessons Learned Library that extracts recommendations and insights from historic MOD Gateway Review reports and turns them into a structured, searchable knowledge base. The solution combines document parsing, NLP categorisation, AI summarisation, and Power BI reporting to surface relevant lessons at project start-up and review stages.
Please be aware that this content was generated follwing an automated review so may not be perfectly accurate; refer to the original challenge brief and team files for authoritative information
Reduced time spent manually reviewing historic Gateway reports; improved reuse of assurance lessons across projects; earlier identification of recurring delivery issues; stronger organisational learning and reduced repeat failures.
33rd times a charm.ipynb: Core notebook for extracting, categorising, and summarising lessons from Gateway Review PDFs.99_problems.py: Automated pipeline to ingest Gateway reports, classify lessons, and append them to a standard Lessons Learned Excel template.20251021-Story Mapping_Lessons Learned Library.docx: Story map outlining the end-to-end Lessons Learned Library workflow and user journeys.
team: Ghost Busters members: Seyi, Ben, Abi topics: solution-centre, hack26, challenge4, python, pdfplumber, sentence-transformers, hugging-face, power-bi, openpyxl, watchdog, lessons-learned, project-assurance, knowledge-management, document-analysis, search, governance, data-quality technologies: python, pdfplumber, sentence-transformers, hugging-face, power-bi, openpyxl, watchdog