Personal observatory · Mexico City

Eduardo Rubén Bernal Puente

Junior Data Analyst at Mercado Libre, analytics engineering for Mercado Pago Point. Data Science at ITAM, class of 2028.

I keep a field sales channel's data correct, on time and usable by the people who run it.

Off the clock I'm Rubo: a music nerd, a gamer and someone who will lose an evening to a telescope. This universe is the informal one.

Enter the observatory View CV

Illustrated portraits · generated from a photo of the author

Mission clocks

At Mercado Libre
------------
since 2025-11-03
Studying Data Science at ITAM
------------
since 2023-01-09
Orbits around the Sun
------------
since 2003-06-15
Until graduation
------------
until 2028-05-10

NGC 3372 · Carina Nebula · Carina · 8,500 ly
M8 · Lagoon Nebula · Sagittarius · 4,100 ly
NGC 7293 · Helix Nebula · Aquarius · 655 ly
M42 · Orion Nebula · Orion · 1,344 ly
M16 · Eagle Nebula · Pillars of Creation · Serpens · 5,700 ly
Barnard 33 · Horsehead Nebula · Orion · 1,375 ly

Observatory

Each nebula is a region of my work. Each star inside it is a fact about a project. Select a nebula to focus the telescope.

The informal half. Each nebula is something I do when I am not working. Pick one.

Select a nebula →

Select a nebula →

Fun facts

The stars in the hero are real: they are computed from a bright-star catalogue for the sky above Mexico City at the moment you open the page.
Fonts, scripts and images are all self-hosted. The only outside request is an anonymous, cookieless page count.
I built my own offline voice-dictation tool (WhisperFlow) because I wanted push-to-talk that never leaves my machine. I dictated the answers that shaped this page with it.

GitHub activity

@rubo6 ↗

215 contributions · last 12 months

Trajectory

Roles, leadership and studies as orbits. The innermost orbit is the most recent.

WorkNov 2025 → presentMexico City
Junior Data Analyst (Data & Analytics Engineering), Contractor
Mercado Libre (Mercado Pago Point) ↗
Acquisition team, Longtail area of Mercado Pago Point (small and mid-size merchants). Since January 2026 I am the data owner for Rangers in Mexico, the field-agent channel that launched here in September 2025. Remote, Spanish-speaking regional team.
- Own the Rangers data layer in Mexico: ETL/ELT pipelines on BigQuery scheduled by an internal orchestrator, about 60 tables (30 in the new backend), daily loads in the millions of rows, five live dashboards in Looker Studio and Google Sheets and five more in progress.
- Rebuilt the agent-to-merchant assignment process. Twelve manual steps across several environments are now one parameterised job: edit a sheet, run it, and the file is ready in under two hours instead of a full working day. Agents, quotas and activations change without touching code.
- Built the internal web platform (Next.js, Firebase, BigQuery, Sheets) where field agents see their own portfolio for the first time. Average monthly subscriptions per agent went from 4 in January 2026 to 18 in June, the month it launched, and 27 by September. Recognised by the area's leadership.
- Own data quality and payouts: I define the data behind the agents' variable compensation, add audit steps that stop a job before an error spreads, answer alert e-mails the same day and run anomaly checks. Two cases of irregular behaviour (May and June 2026) were flagged for investigation this way.
- Automated the monthly regulatory report to Mexico City's mobility authority on transit-card top-ups at Point terminals. Hand-run queries are now five self-refreshing dashboards.
- Production Python (pandas, pyarrow, google-cloud-bigquery) with pytest, and Git with code review.
- Weekly planning and one-to-ones with operations, invoicing, payouts, cashback and measurement stakeholders; monthly and quarterly business reviews.
- Mentor two colleagues on the team's data tooling.
- BigQuery
- SQL
- Python
- Looker Studio
- Google Sheets
- Next.js
- Firebase
- pytest
- Git
LeadershipJan 2025 → Dec 2025Mexico City
Research Manager, ECOBOX AI team member
DataLab ITAM
- Member of the ECOBOX AI team, a computer-vision model for waste classification: the team delivered the trained model and a paper (linked from my LinkedIn); the project did not reach a deployed prototype.
- Research planning and technical documentation for the student data-science community.
- Applied Python, machine learning and data analysis to a project with environmental impact.
- Python
- Computer Vision
- Research documentation
EducationNov 2025 → Jul 2026Mexico City
Artificial Intelligence Diploma Program
Universidad Iberoamericana ↗
- Continuing Education 120-hour programme: foundations and paradigms of AI, machine learning algorithms, supervised and unsupervised learning in Python, intelligent agents and chatbots, RPA, natural language processing and artificial neural networks, including AI ethics.
- Final project: a chatbot with an n8n automation, built with the Python machine-learning stack covered in the programme.
- Diploma and certificate expected in December 2026.
- Machine Learning
- Deep Learning
- Chatbots
- n8n
About the institution and full coursework
- Reputation: Jesuit university founded in 1943, member of the Jesuit University System (SUJ) and AUSJAL, about 12,000 students. 851–900 worldwide and 6th in Mexico in the QS World University Rankings 2027; institutional CIEES accreditation 2025–2032; first in Design and second in Architecture and Accounting in Reforma's 2025 employer survey.
Sources:QS 2027, Mexico (Vanguardia) ↗IBERO · Reforma 2025 ↗IBERO · accreditations ↗IBERO · AI diploma programme ↗
LeadershipJul 2024 → Dec 2024Mexico City
Team Leader, Outgoing Global Volunteer
AIESEC ITAM ↗
- Led a four-member team across the attraction-to-realization pipeline, setting priorities, tracking KPIs and running recurring 1:1 follow-ups.
- Coordinated exchanges with Italy, Colombia, Peru and other Mexican states; communication with international stakeholders in English.
- Supported nine international volunteer realizations and improved follow-up across the participant pipeline.
- Attended AIESEC regional and national conferences on the organisation’s leadership culture.
- Leadership
- Stakeholder communication
LeadershipJan 2024 → Jun 2024Mexico City
Team Member, Incoming Global Volunteer
AIESEC ITAM ↗
- Supported eight international volunteers from pre-arrival through local project execution and follow-up.
- Coordinated with local NGOs to align projects with the Sustainable Development Goals, logistics and participant safety.
- Strengthened intercultural communication, stakeholder coordination and problem-solving in dynamic environments.
- Intercultural communication
- NGO coordination
EducationJan 2023 → May 2028Mexico City
B.S. in Data Science
Instituto Tecnológico Autónomo de México (ITAM) ↗
Expected graduation: May 2028 (nine-term plan; electives to be taken inside a master’s programme so both degrees finish together).
- Relevant coursework: Machine Learning, Artificial Intelligence, Databases and NoSQL Databases, Parallel and Cloud Computing, Big Data Architecture, Data Mining, Bayesian Statistics, Causal Inference. Tools by course: R, Java, Python, Bash, C++ with OpenMP and MPI, MongoDB, Cassandra, Neo4j, Spark, Docker.
- Academic focus: data engineering, machine learning, computer vision and parallel computing.
- Favourite courses so far: Data Structures, Artificial Intelligence, Data Sources and, this term, Architecture for Large Data Volumes. Final projects included a website built from scratch (Data Sources) and a machine-learning goal-probability model with a game-like interface (Artificial Intelligence).
- Tutor for Probability I; planning a terminal project rather than a thesis.
- Statistics
- Algorithms
- Parallel Computing
- NoSQL
About the institution and full coursework
- Reputation: 4th university in Mexico and 751–760 worldwide in the QS World University Rankings 2027, and 1st in Mexico / 30th worldwide in the QS Employment Outcomes indicator (95.5 % of graduates employed within 7–11 months). 42nd in QS Latin America 2026. Economics is ranked =185 worldwide (QS by Subject 2025) and first in Reforma's employer survey in 2024, 2025 and 2026; Computer Science entered QS by Subject in 2025 (701–750).
- The Data Science degree (8 semesters, Statistics Department founded in 1997) reports that 100 % of its 2020–2022 graduates are working or in graduate school, with an average of six job offers during the programme. Full FIMPES accreditation 2023–2030.
- Plan B coursework (official syllabi): Calculus I–III, Vector Geometry, Linear Algebra, Mathematical Thinking, Discrete Mathematics; Probability I–II, Mathematical Statistics, Linear Methods (GLMs), Multivariate Methods, Bayesian Statistics; Algorithms and Programs, Data Structures (Advanced), Databases, NoSQL Databases, Data Sources, Information Visualization, Machine Learning, Artificial Intelligence, Parallel and Cloud Computing, Big Data Architecture, Data Mining; Economics I–IV, Market Design, Causal Inference; Legality and Ethics in Data Science; and the ITAM humanities core (Ideas and Political Institutions I–III, Contemporary Civilization, Mexican history and reality).
- Tooling by course: R across the statistics sequence; Java in the data-structures sequence; Python, Bash, Linux, regex and Git in Data Sources; MongoDB, Cassandra, Neo4j, Spark/Hadoop and Docker in NoSQL; C++ with OpenMP and MPI in Parallel Computing; Python agents and search in AI.
Sources:QS 2027, Mexico (Vanguardia) ↗ITAM · QS Employment Outcomes 2027 ↗ITAM · QS by Subject 2025 ↗ITAM · Reforma 2026 ↗ITAM · Data Science degree ↗ITAM · Plan B bulletin (CDA-B) ↗ITAM · statistics syllabi ↗ITAM · Machine Learning syllabus ↗
EducationAug 2018 → Jul 2021Mexico City
Technical Degree in Programming (Técnico en Programación)
CECyT No. 9 "Juan de Dios Bátiz" (Instituto Politécnico Nacional) ↗
IPN bivalent upper-secondary programme (high-school diploma + technical degree), Plan 2008. Title issued December 2022.
- Programming track: object-oriented and web programming, database design, test engineering (JUnit), web and application security, distributed systems, agile methods (Scrum, XP) and four semesters of IT project labs.
- Lab projects: Java desktop apps (a calculator, an inventory and cargo manager for a shipping company), hand-coded HTML5 sites, MySQL database design and exercises. Final project: a 2D web videogame in PHP where kids solve chemistry exercises to advance levels. First language: Java; then C/C++, PHP, JavaScript, Python and SQL.
- Admitted with 118 of 128 correct answers in the 2018 COMIPEMS exam (the campus cut-off that year was around 105). First place at a school fair with an environmental-awareness project.
- Graduated in July 2021 after a year and a half of online classes during the pandemic (remote exams were made deliberately hard); the technical title (Técnico en Programación) was issued in December 2022 because the titling exam had to wait for in-person sessions.
- C++
- Java
- JavaScript
- SQL
- Software testing
About the institution and full coursework
- The Bátiz is the IPN campus with the highest COMIPEMS cut-off score in the country, roughly ten points above the next one.
- Selectivity: the CECyT 9 cut-off in the COMIPEMS entrance exam was 102–106 of 128 questions every year between 2019 and 2023, the highest of any IPN campus and about ten points above the second; the press calls it the hardest IPN campus to get into.
- Ranking: in the 2024 COMIPEMS the CECyT 9 cut-off was 102 of 128, the highest of any IPN campus (the next was 94) and above UNAM's CCH Sur; the top scorer of the whole 2024 exam (126/128) chose the Bátiz. It was named Mexico's best high school in the ENLACE test five years in a row up to 2012 (92 % of students at the excellent level in Mathematics) and in 2015 it was still the only public high school in the national top 5 both by admission demand and by university admission rate.
- Programming track: Object-Oriented Programming, Web Programming and Services, Databases (normalized design and data manipulation languages), Introduction to Test Engineering and Test Automation, Web and Application Security, Introduction to Distributed Systems, Agile Programming Methods, Basic Software Engineering, and four semesters of IT Project Labs.
- Standards and tools named in the official Plan 2008 programmes: UML (all diagram types), JUnit and white/black-box test design, XP, Scrum and Crystal, SOAP/WSDL/UDDI web services, sockets and RPC, symmetric and public-key cryptography, SSL/TLS, Kerberos and IPsec, PSP, CMMI and COCOMO, MoProSoft.
- Full scientific core alongside the technical track: Algebra, Geometry and Trigonometry, Analytic Geometry, Differential and Integral Calculus, Probability and Statistics, Physics I–IV, Chemistry I–IV, plus philosophy, history and six semesters of English.
Sources:CECyT 9 · Plan 2008 programme (official PDF) ↗CECyT 9 · syllabi index ↗El Universal 2012 · ENLACE ↗Excélsior 2015 ↗COMIPEMS 2024 cut-offs ↗Milenio · top scorer 2024 ↗

Skills

Grouped as constellations. Brighter stars are stronger skills.

Hover or focus a star to see the course or job behind it.

Analytics & ML

Scorpius

Data & Cloud

Ursa Major

Engineering

Cygnus

Mathematical foundations

Lyra

Languages

Orion

Soft skills, with evidence

Team leadershipLed a four-member AIESEC team through the attraction-to-realization pipeline (nine international volunteers placed) and mentor two colleagues on the data tooling at Mercado Libre.
Stakeholder communicationWeekly planning and one-to-ones with operations, invoicing, payouts, cashback and measurement teams; monthly and quarterly business reviews; earlier, international coordination in English at AIESEC.
Documentation and knowledge transferArchitecture notes, changelogs and handoffs kept current so the team can operate and extend what I build without me.
Ownership under ambiguityTook over the Rangers data layer two months after joining and rebuilt a twelve-step manual process into one parameterised job while the backend was being migrated.
Data quality mindsetAudit steps that stop a job before an error spreads, same-day response to alerts, anomaly checks that surfaced two irregular cases for investigation.
Research coordinationResearch Manager at DataLab ITAM: planned and documented the ECOBOX AI work so a student team with exams could still deliver a model and a paper.
Long-term disciplineFifteen years of swimming and more than a decade of piano alongside a full course load and a job.

Certifications in progress

In progressAnthropic certification (in progress)Anthropic
In progressAWS Academy — Cloud Foundations, Data Engineering, Intermediate (in progress)AWS Academy
In progressDataCamp — Data Engineer track incl. Git (in progress)DataCamp
ExpectedArtificial Intelligence Diploma ProgramUniversidad Iberoamericana · 2026-12
EarnedTechnical Degree in Programming (título de Técnico en Programación) ↗Instituto Politécnico Nacional · CECyT No. 9 · 2022-12

Observation log

A study log of the B.S. in Data Science at ITAM, one entry per term: the mathematics and statistics underneath, the computing, the economics, and the humanities the programme insists on. Written as I go, in the first person.

All entries →

Contact

Open to data and analytics engineering roles. Email gets the fastest reply.

[email protected]

GitHub

github.com/rubo6 ↗

linkedin.com/in/eruben-datascience ↗

Spotify

open.spotify.com/user/d6hedn5s0vafnjx1co5uz8gdb ↗

Xbox

xbox.com/play/user/Mthra ↗

Languages

Spanish · Native
English · Professional working proficiency

Eduardo Rubén Bernal PuenteRubo

Mission clocks

Parallel Bag-of-Words

ITAM coursework repositories

WhisperFlow — offline voice dictation for Windows

This observatory

Internal data platform for Mercado Pago Point

ETL/ELT pipelines and warehouse models for Point

Keeper Save Probability

ECOBOX AI

Fun facts

GitHub activity

Junior Data Analyst (Data & Analytics Engineering), Contractor

Research Manager, ECOBOX AI team member

Artificial Intelligence Diploma Program

Team Leader, Outgoing Global Volunteer

Team Member, Incoming Global Volunteer

B.S. in Data Science

Technical Degree in Programming (Técnico en Programación)

Soft skills, with evidence

Certifications in progress

From syllabi to stars: how I mapped my skills to their sources

WhisperFlow: why I built voice dictation that never leaves my machine

Pulsars: the lighthouses that keep time better than an atomic clock

Eduardo Rubén Bernal Puente