CodeandCoffeeCommunity / speaking

13 stars 3 forks source link

How I built data pipeline as a One-Person Data Team in early stage startup #12

Closed dq-hustlecoding closed 1 year ago

dq-hustlecoding commented 1 year ago

About You

Your Name: DQ Choi

Twitter handle (optional):

The best way to reach out to you: dq.choi@momenti.tv

A quick bio: 8+ years experienced data scientist/engineer data engineer at @corp-momenti (see our homepage here) Worked for various early stage startups author of "Build Your Own Netflix"

Your Talk

Title: How I Built a Data Pipeline from Scratch as a One-Person Data Team: A Practical Guide

What your talk is about: A year ago, I joined Momenti, a new media platform startup, as the sole member of the data team. The existing data pipelines were built on GCP Spanner and Elixir, which were decent but too costly for an early-stage startup with a limited budget. Therefore, I decided to create a new data pipeline from scratch using GCP Pubsub and Bigquery and NestJS server. As a result, we were able to reduce our GCP usage fee by up to 90% compared to the previous pipeline. In this talk, I will share the insights and challenges I gained from this journey, as well as the areas for further improvement.

How long will your talk be?

Meta

Do you need help crafting your talk?

Meetup event Copy - Optional

Event Title:

Event Desciption:

Danamitecoder commented 1 year ago

Hi DQ Choi! Thank you so much for submitting a request to give a talk at a future Code and Coffee. I'll email you directly so we can touch base regarding your talk and next steps!