M MOCHO
Independent · Student-led · Early stage

Research, data and AI built to be useful.

Mocho is an independent research and technology initiative working at the intersection of statistics, data science and artificial intelligence — from official statistics and health-economics research to accessible tools for education and public benefit.

Explore our focus View GitHub ↗
Stage Early-stage initiative
Approach Independent & bootstrapped
Purpose Research & public benefit

About

Small by design. Serious about the work.

Mocho is a student-led independent initiative founded by Moises Ignacio Villena Mejia, a Statistics student at ESPOL. It exists as a space to develop research ideas, experimental tools, reproducible workflows and practical applications of AI and data science.

The initiative is currently independent and early-stage. Its goal is to turn rigorous technical work into tools and resources that are understandable, useful and accessible to more people.

Disclosure: Mocho is an independent initiative and is not affiliated with or endorsed by ESPOL.

Selected work

Applied statistics with real public-data problems.

Selected statistical work developed for Ecuador's National Institute of Statistics and Censuses (INEC), focused on the Health Satellite Accounts and on improving the continuity and timeliness of official health-sector information.

2026 · Official statistics Research project

Health Satellite Accounts Backcasting

Reconstruction of Production, Intermediate Consumption and Gross Value Added series for 2003–2006.

Developed a backcasting workflow to make historical Health Satellite Account series compatible with the current methodological framework. The work combined classification harmonization, value indices, aggregation, rescaling and redistribution procedures to preserve accounting consistency across methodological changes.

Backcasting Official Statistics Health Economics Index Methods Data Harmonization
Why it matters: creates a more coherent historical series for long-run analysis of Ecuador's health-sector production and value added.
2026 · Time series Forecasting project

Out-of-Pocket Health Expenditure Forecasting

ARIMAX-based forecasting of household out-of-pocket health expenditure for 2025–2029.

Built a time-series forecasting workflow using Health Satellite Account data from 2007–2024 and public health expenditure as an exogenous variable. The analysis included stationarity assessment, model selection, residual diagnostics, ex-post validation and uncertainty intervals.

ARIMAX Time Series Forecasting Model Validation Health Economics
Why it matters: explores how statistical forecasting can partially reduce the lag between real-world conditions and the availability of official health indicators.
Project descriptions summarize research reports authored by Moises Villena Mejia. References to INEC describe the institutional project context and the official statistical framework used in the work; they do not imply that Mocho is an official INEC program or product.

Focus areas

Where Mocho is exploring.

Work is centered on areas where statistical reasoning, software and AI can improve how people learn, investigate questions and turn data into decisions.

01

Applied Statistics & Data Science

Reproducible analysis, statistical modeling, visualization and practical workflows for working with real data.

02

AI-Assisted Research

Exploring responsible uses of language models to support research, synthesis, prototyping and technical workflows.

03

Educational Technology

Tools and experiments designed to make difficult technical concepts easier to study, explain and apply.

04

Open & Accessible Tools

Small, practical software and resources with an emphasis on accessibility, transparency and public usefulness.

Principles

Useful technology starts with good standards.

Mocho is not built around hype. The goal is to make careful, testable work that people can understand and reuse.

Build things that are clear enough to inspect, rigorous enough to trust and useful enough to matter.

✓
Accuracy Prefer evidence, validation and explicit limitations.
✓
Reproducibility Make methods and assumptions understandable wherever possible.
✓
Accessibility Reduce unnecessary barriers to learning and using technical tools.
✓
Public benefit Favor work that can create practical value beyond the project itself.

Contact

Interested in research, collaboration or what Mocho is building?

For research, collaboration or project inquiries, reach Mocho directly through its domain email or follow the work on GitHub.

GitHub ↗ moises@mocho.dev