Skip to content
Sample profile. This is an example profile that shows how Introworth works, not a real person. You can send a message to try the flow, but nobody reads or replies to it and you are never charged.

Juliana Valencia

Sample profile

Staff Data Engineer at Delta Systems · Kafka, Airflow

Works at
Delta Systems
Location
Mexico City, Mexico · CST (UTC-6)
Experience
9+ years experience
Education
Unicamp
Languages
English, Spanish

Overview

I've spent 9 years on data pipelines - first at América Móvil, now at Delta Systems. Shipped a rewrite that took build times from 32 minutes to under 3. Wrote the Snowflake service that now handles 62 million requests a day. Most at home with SQL and Airflow. Coffee nerd.

Good reasons to reach out
  • Working sessions on Snowflake
  • Second opinions on Spark decisions
  • Questions about demand forecasting
Please don't message about
  • SEO and link-building offers
  • Survey or panel invitations
  • Generic networking requests

Best reached early in the week

Skills
KafkaIcebergSQLAirflowSnowflakeSpark

Experience

  1. Staff Data Engineer
    Apr 2026 - Present
    Delta Systems · 7 mo · Full-time · Mexico City, Mexico
    • Reduced stock-outs by 34% in the first season.
    • Built the forecasting models now used in 3 product decisions a quarter.
    • Wrote the metric definitions the leadership team reviews every week.
  2. Senior Data Engineer
    Sep 2023 - Apr 2026
    McKinsey & Company · 2 yr 7 mo · Full-time · Mexico City, Mexico
    • Replaced 45 hand-built dashboards with one metrics layer people trust.
  3. Senior Data Engineer
    Jan 2021 - Sep 2023
    Cemex · 2 yr 8 mo · Full-time · Mexico City, Mexico
    • Taught SQL to 200 non-analysts.
  4. Data Engineer
    Sep 2019 - Nov 2020
    Vale · 1 yr 2 mo · Full-time · Mexico City, Mexico
    • Taught SQL to 180 non-analysts.
  5. Data Engineer
    Dec 2017 - Sep 2019
    América Móvil · 1 yr 9 mo · Full-time · Mexico City, Mexico
    • Replaced 22 hand-built dashboards with one metrics layer people trust.
    • Cut forecast error from 32% to 8% across 1270 SKUs.

Projects

Conference talk: lessons from forecasting
2025

What went wrong and what worked while scaling demand forecasting.

Open-source Spark toolkit
2024

A small library for demand forecasting that 375 developers star on GitHub.

References

No references yet.

Education

  1. Unicamp
    Bacharelado, Computer Science
    2012 - 2016

Details

Compensation expectations
What they are targeting for their next role.
Visible to connections
Direct contact details
Direct email and a scheduling link.
Visible to connections