Description of Organization
The Collective Energy Foundation (FEC) is a Canadian charitable organization focused on advancing education and innovation. We collaborate with communities, partners, and volunteers to create meaningful impact through technology, sustainable development, and environmental stewardship. From soil to software, we believe sustainable living, wellness, and innovation should be within reach for everyone. Through our education model, we empower students, researchers, and communities to build solutions from the ground up, from regenerative growing practices to advanced technologies like AI and large language models.
Problem area
FEC's organizational content (mission statements, project details, historical records, technical documentation) is fragmented across documents and platforms, forcing staff to manually recompile and reformat it every time they need materials for donors, sponsors, or legal bodies. Students will explore how to consolidate this into a single, internal source of truth.
Main objectives
Design a structured content tagging and management system for internal FEC staff use, including import support for existing Word/PDF documents. Build a transformation engine, using LLMs where appropriate, that converts tagged content into one target output format (for example, a donor one-pager or slide deck) against defined accuracy and consistency criteria. Deliver a one-click generation interface so non-technical staff can produce that output without manual reformatting. Stretch goals: extend to additional output formats (compliance documents, fundraising brochures) and add content freshness or usage analytics.
Scope of work
Audit FEC's existing content sources and formats; design the content tagging schema and repository structure; build the import pipeline for legacy Word/PDF documents; develop the LLM-based transformation engine for the chosen target output; build the one-click generation interface; test output accuracy and consistency against defined criteria with real FEC content.
Deliverables
- Report
- Presentation
- Website
- Resources
Frequency
- Weekly
Skills and training required
- Data modeling and content architecture
- LLM/prompt engineering
- Full-stack web development (React/TypeScript front end, Express.js back end)
- Document parsing (Word/PDF)
- Interface design for non-technical end users
Resources required
- Access to FEC's existing organizational documents and content sources
- An LLM API or hosting environment
- The existing tech stack (React/TypeScript, Express.js, Keycloak, PostgreSQL)
- Staff input to define what counts as accurate on-brand output for the target format
NDA or a commercialization agreement for this project?
Yes