Nicolás Saldarriaga – Welcome Center

Desktop

Nicolás Saldarriaga Carvajal

Data and automation for development finance and food systems

I turn the information institutions sit on into systems they can decide with. I have done it at the IDB Treasury, in IFAD’s Latin America and Caribbean division and at the OECD. Today I build and run two production platforms for Prisier, a pricing consultancy in Bogotá.

101,156IDB Treasury files restructured into 5 categories
US$576MIFAD financing across the 33 projects I helped map
428Commits across both Prisier platforms, Jul–Oct 2026
11 daysPrisier One, from first commit to production
891Automated tests passing in Prisier One
15 yearsOf FINAGRO credit data analysed, 2010–2024

Where finance, data and software overlap

Four facets, one profile. Pick one to see where it sits and what it touches; the names in the sentences light the overlaps.

the field Food systems the setting Development finance the method Applied AI Agrifood policy & investment Computational food research Structure for institutional data Evidence systems

Food systems

The field

Trained in
Food Engineering, Universidad de los Andes, with a minor in Development Studies
The questions
yields, losses and costs along the chain, and who carries them
Where
OECD work on food loss and waste; Uniandes research on post-conflict territories

Meets development finance in and applied AI in .

Development finance

The setting

Institutions
IDB Treasury in Washington, IFAD's Latin America and Caribbean division in Rome
Today
Prisier, a pricing consultancy in Bogotá: its treasury model and two production platforms
The habit
decisions documented, numbers traceable to their source

Meets food systems in and applied AI in .

Applied AI

The method

In production
Prisier BOps and Prisier One, platforms in daily use
Tools
Python, K-means and TF-IDF, BM25 retrieval, rule-based classification, LLM applications
The rule
a model earns its place only when it changes a decision

Meets food systems in and development finance in .

Evidence systems

What I deliver

Evidence systems: the place where the three circles meet.

Meaning
the messy information an institution sits on, turned into something it can decide with
Shape
a structure, a pipeline and a page that stay useful after I leave

Every overlap on the diagram is one of these: , , .

The full picture

The full picture

I turn the messy information institutions sit on into evidence they can decide with. I learned the questions inside IFAD, the IDB and the OECD, and I build the systems, now with AI, that answer them.

IFAD, 33 projects in Latin America and the Caribbean, US$576M; OECD background notes
Uniandes, K-means and TF-IDF over 33,000+ development records
IDB Treasury, 101,156 investment files into 5 categories and 24 subfolders

Bogotá, ColombiaUpdated October 2026

Bogotá time