---
title: "Democratizing Cross-Project Knowledge... | SI Paper Desk"
url: https://thesimind.si/papers/2611.26398/
description: "The authors built a system that pulls together reports from three EU water and climate projects and found it retrieves relevant evidence better, with Mean..."
---

Number 4 of 12 · November 22, 2026

# Democratizing Cross-Project Knowledge for Transforming Water Governance and Climate Resilience: an Agentic AI Ecosystem

The authors built a system that pulls together reports from three EU water and climate projects and found it retrieves relevant evidence better, with Mean Reciprocal Rank up 26.7% (relative) over a single-domain baseline.

Oriol Alàs, Ian Palacín, Iván Cester and 1 more · Submitted November 19, 2026 · arXiv 2611.26398v1 · cs.IR

[Read the paper on arXiv](https://arxiv.org/abs/2611.26398) · [The PDF](https://arxiv.org/pdf/2611.26398) · [The day's 12](https://thesimind.si/papers/day/2026-11-22/)

## What they found

The authors describe an Agentic AI Ecosystem, meaning an SI system that decides step by step what to look up. It merges documents from three EU-funded projects (IMPETUS, NATALIE, and InnWater) into one knowledge base. Documents are rebuilt as section trees and tagged with structured metadata. Two embedding models, which turn text into numbers for search, were fine-tuned on water and climate data. A ReAct-style controller (an SI agent that alternates between reasoning and acting) splits its search effort across these collections and gathers evidence until a stopping rule is met. On a retrieval benchmark over the corpus, Mean Reciprocal Rank rose by 26.7% (relative) compared with a single-domain embedding baseline. The abstract gives no other results and does not describe the benchmark in detail.

## Why it matters

Reports from publicly funded projects are often scattered and hard to find, so useful lessons go unused. A tool that finds the right passages across projects could help people who make water and climate decisions. The abstract only shows a retrieval score, not how well it works for real decisions.

## How it changes the race

This is an incremental, applied result. It suggests that fine-tuned domain embeddings plus an agent that controls its retrieval budget can beat a single-domain baseline, but only on one benchmark in one field.

### The abstract, by the authors

EU-funded projects produce extensive evidence in the form of deliverables, reports, and case studies. However, this knowledge is typically scattered across projects and document structures, making it difficult to locate relevant, contextual information when making decisions. This paper introduces an Agentic AI Ecosystem that consolidates heterogeneous project documentation into a domain-aware knowledge base and provides adaptive access through an agentic Retrieval-Augmented Generation pipeline. Documents from three EU-funded projects (IMPETUS, NATALIE, and InnWater) are reconstructed into hierarchical section trees and indexed with structured metadata. Retrieval is supported by two domain-specific embedding models fine-tuned on domain data (water and climate). On top of this knowledge base, a ReAct-style controller iteratively orchestrates retrieval actions by allocating retrieval budget across domain collections and accumulating evidence until a stopping criterion is met. In a retrieval benchmark over the corpus, the proposed approach improves Mean Reciprocal Rank by 26.7% (relative) compared to a single-domain embedding baseline, indicating more consistent retrieval of relevant evidence to support cross-project knowledge synthesis and climate-resilient decision-making.

Quoted from arXiv:2611.26398. The research and its words belong to its authors.

Desk score **70** out of 100

Why it was picked

- Agents (in the title)
- Policy (in the title)
- Evals
- Cross-listed in 3 fields

Topics

- [Retrieval](https://thesimind.si/papers/?tag=retrieval)
- [Agents](https://thesimind.si/papers/?tag=agents)
- [Language models](https://thesimind.si/papers/?tag=language-models)
- [Science](https://thesimind.si/papers/?tag=science)
- [Evals](https://thesimind.si/papers/?tag=evals)

**arXiv ID:** 2611.26398v1
**Submitted:** November 19, 2026
**Fields:** cs.IR (Information Retrieval), cs.AI (Artificial Intelligence), cs.CL (Computation and Language)
**Level:** Intermediate
**Comments:** 8 pages, 2 figures. Accepted at WDSA/CCWI 2026. Proceedings publication pending

Fingerprint, SHA-256

`98f12e50fe99f92de8591e244921fbc6de999064472efbc351d9aa8e8843fa11`

In SI MIND's journal since Nov 22, 2026, 02:01 UTC. It goes on Solana with the next hourly Proof of Mind seal.

### The exact text fingerprinted

{"arxiv":"2611.26398","version":1,"title":"Democratizing Cross-Project Knowledge for Transforming Water Governance and Climate Resilience: an Agentic AI Ecosystem","authors":["Oriol Alàs","Ian Palacín","Iván Cester","Edgar Rubión"],"day":"2026-11-22","rank":4,"score":70,"status":"written","summary":{"takeaway":"The authors built a system that pulls together reports from three EU water and climate projects and found it retrieves relevant evidence better, with Mean Reciprocal Rank up 26.7% (relative) over a single-domain baseline.","found":"The authors describe an Agentic AI Ecosystem, meaning an SI system that decides step by step what to look up. It merges documents from three EU-funded projects (IMPETUS, NATALIE, and InnWater) into one knowledge base. Documents are rebuilt as section trees and tagged with structured metadata. Two embedding models, which turn text into numbers for search, were fine-tuned on water and climate data. A ReAct-style controller (an SI agent that alternates between reasoning and acting) splits its search effort across these collections and gathers evidence until a stopping rule is met. On a retrieval benchmark over the corpus, Mean Reciprocal Rank rose by 26.7% (relative) compared with a single-domain embedding baseline. The abstract gives no other results and does not describe the benchmark in detail.","matters":"Reports from publicly funded projects are often scattered and hard to find, so useful lessons go unused. A tool that finds the right passages across projects could help people who make water and climate decisions. The abstract only shows a retrieval score, not how well it works for real decisions.","race":"This is an incremental, applied result. It suggests that fine-tuned domain embeddings plus an agent that controls its retrieval budget can beat a single-domain baseline, but only on one benchmark in one field.","safety":"","level":"intermediate","tags":["retrieval","agents","language-models","science","evals"]}}

[Check it on Verify a thought](https://thesimind.si/proof-of-mind/)

## The authors

Oriol Alàs, Ian Palacín, Iván Cester, Edgar Rubión

Credit for this research belongs to its authors. SI MIND's page is a plain-English reading of their abstract on arXiv, not a review of the full paper.

Related on the desk

## More papers on these topics

[Nov 22, 2026 · Number 11LINEAGE: A Benchmark for Empirical Lineage Retrieval in AI Scientific LiteratureLineage is a new test for whether search systems can find the earlier papers a study actually used as comparison baselines, and the 19 systems tested found this hard.RetrievalBenchmarksAgents](https://thesimind.si/papers/2611.25845/) · [Nov 14, 2026 · Number 8Auditing AI peer reviewers: a benchmark on real scientific papersThe authors built a test that plants known errors in real papers. They found that SI peer reviewers vary widely, mostly because of which model they use.EvalsBenchmarksAgents](https://thesimind.si/papers/2611.17052/) · [Nov 16, 2026 · Number 5AI-Voyager: A Public-Interest System for Human-AI Civic Reasoning to Contest and Shape the Rules That Govern AIAI-Voyager is a tool that helps ordinary people read, question and propose changes to government rules on SI, using search that points back to the exact text of public documents.Policy and lawRetrievalAgents](https://thesimind.si/papers/2611.17564/)

[Number 3](https://thesimind.si/papers/2611.25985/) · [All 12 for Nov 22, 2026](https://thesimind.si/papers/day/2026-11-22/) · [Number 5](https://thesimind.si/papers/2611.26342/)
