ACCEM
AI-Powered RAG System for Program Analysis & Service Matching
Third Sector / Non-profit
|Ongoing — continuous consulting35%
Effectiveness Increase
10x
Faster Matching
1000s
Documents Indexed
The Challenge
ACCEM runs dozens of social programs across Spain serving refugees, migrants, and vulnerable populations. Program managers struggled to efficiently match beneficiaries with the right services, analyze program effectiveness across regions, and extract actionable insights from thousands of case files and reports scattered across multiple systems.
Our Approach
We designed and deployed a Retrieval-Augmented Generation (RAG) system that ingests ACCEM's program documentation, case files, and operational reports into a vector store for semantic search. The AI layer enables program managers to query their data in natural language — finding relevant precedents, matching beneficiaries to available programs, and generating summary reports. The system integrates with their existing PostgreSQL infrastructure and respects strict data access controls to protect sensitive beneficiary information. We provide ongoing consulting to refine the model, expand the knowledge base, and develop new AI-powered workflows as ACCEM's needs evolve.
Key Deliverables
Tech Stack
Impact
The RAG system transformed how ACCEM's program managers access and use their institutional knowledge. Beneficiary-to-program matching time dropped from hours of manual review to seconds of AI-assisted search. Program effectiveness analysis that previously required weeks of manual data compilation can now be generated on demand. The system increased overall program effectiveness by 35%, helping ACCEM serve more people with better-targeted resources. We continue as embedded AI consultants, expanding the system's capabilities and training staff on new AI-powered workflows.
Interested in similar results?
Let's talk about your project and how we can help.