# Julio Alberto Guevara Nuñez

Julio Alberto Guevara Nuñez is a remote worker on Himalayas, ML Engineer at Cartodata, based in Mexico.

Source: https://himalayas.app/@julioalbertoguevaran

```text
Machine Learning Engineer at Cartodata improving computer vision models for geospatial imagery.
```

| Detail | Value |
| --- | --- |
| Availability | Looking for a job |
| Location | Mexico |
| Authorized to work in | Mexico |
| Salary expectations | 207k+ MXN |
| Job categories | Machine Learning Engineer, GIS Analyst |
| Skills | Machine Learning, Computer Vision, Remote Sensing, GIS, Object Detection, Geospatial Evaluation, Data Labeling, MLOps, Versioned Datasets, Reproducible ML, Optical Character Recognition, PyTorch, MMDetection, RF DETR, Faster RCNN |
| LinkedIn | https://linkedin.com/in/julio-alberto-guevara-nunez-886885249 |
| GitHub | https://github.com/Julio93800 |

## What I'm looking for

```text
I’m looking for 100% remote, flexible contract work in the evenings (Mexico time), with overlap with U.S. Pacific Time, alongside my main role.
```

## About

```text
At Cartodata, I improved an urban object detection model on 360-degree imagery 5.7x, reaching mAP@[.5:.95] = 0.43 with ResNet50. I used class-balanced sampling, conditional data augmentation, and geospatial evaluation with GeoPandas.

I also built an OCR pipeline with PaddleOCR to classify scanned cadastral records and export them to PDF, and I’m developing a road surface classifier from street-level imagery. I identified and documented a data leakage issue in the training and evaluation workflow.

At Cartodata, I coordinate annotation projects in CVAT and built a performance dashboard in Grafana to track labeling progress. I wrote the team standard for documenting and versioning datasets and models, and track experiments in MLflow.

Before moving into machine learning, I supported municipal cadastral offices in Jalisco as a GIS Analyst, linking current and historical cadastral data to cartography and maintaining cartographic and cadastral databases. My projects include a natural-language GIS agent that operates QGIS, quarry detection on Sentinel-2 imagery, and a service that turns FAA NOTAMs into geographic data.
```

## Experience

### ML Engineer (current)

Cartodata. Jan 2025 - Present

```text
Improved urban object detection model on 360-degree imagery 5.7x to mAP@[.5:.95] = 0.43 through class-balanced sampling and conditional data augmentation. Built OCR pipeline for cadastral records and leads data labeling team.
```

Skills: PyTorch, MMDetection, RF-DETR, Faster-RCNN, YOLO, Paddle-OCR, CVAT, Geopandas, MLFlow, Pixi, Grafana, Clickhouse, PostgreSQL

### GIS Analyst

Cartodata. May 2023 - Dec 2024

```text
Provided cartographic support for municipal cadastral offices in Jalisco, linking current and historical cadastral data to cartography. Updated cadastral information for the Centro Nacional de Registros (CNR) project in El Salvador.
```

Skills: QGIS, PostGIS, Cartography, Cadastral, Database-Management

## Education

### Universidad de Guadalajara

B.Eng., Geophysics. 2018 - 2023

```text
B.Eng. in Geophysics from Universidad de Guadalajara, completed from 2018 to 2023.
```

## Tech stack

- PostGIS
- PostgreSQL
- Python
- PyTorch
- MLflow
- FastAPI
- Grafana
- Pixi
- Linux
- Git
- Docker
- Clickhouse
- LangChain
- LangGraph
- QGIS

To contact Julio Alberto, message them from this profile with a Himalayas employer account: [Hire on Himalayas](https://himalayas.app/hire).

## People also viewed

- [Alan Guerrero](https://himalayas.app/@alanguerrero1): Data Science Area Director, PROFEPA
- [Camilo CovHer](https://himalayas.app/@mxtxtrxn): Machine Learning Researcher, University of Guanajuato
- [Mauricio Gutiérrez Hernández](https://himalayas.app/@mauriciogutirrezhern): Participante Técnico, Incubadora ESCOM Google Cloud y Nebursaturn Academy
- [julio cesar Valderrabano Henandez](https://himalayas.app/@juliocesarvalderraba): Científico de Datos Junior, Freelance
- [Cmaps User](https://himalayas.app/@Jcardena): Independent Geospatial & Territorial Data Specialist, Project Sección C
- [Francisco Javier Urías Ortiz](https://himalayas.app/@franciscojavieruraso): Data Engineer, SINECTA.IA
- [Erick Aguilera](https://himalayas.app/@erickaguilera): Deep Learning Engineer, BeagleMind GmbH
- [María Soto](https://himalayas.app/@marasoto): Variable Cost Productivity Intern, GE Aerospace
- [Rogelio Lizárraga](https://himalayas.app/@rogeliolizrraga): Data Manager, Desarrollo Geomatico
- [Jorge Luis Herrera Saavedra](https://himalayas.app/@jorgeluisherrerasaav): Machine Learning Intern, Oracle
