Se dividio en 3 fases.
La primera una screening con RH en donde se hablo del puesto.
La segunda conociendo al equipo de desarollo y profundizar en tu experiencia.
La tercera un ejercicio de caso de uso
La cuarta defender y explicar todo tu razonamiento al resolver el caso de uso
Interview questions [1]
Question 1
Como aplicarias la taxonomia aun grupo de productos usando LLM? Considera la ingesta y respuesta de los tokens para que sea de la forma mas eficiente
2 Teams calls and 1:1. HR first met with me and recommended that I meet the VP. Great experience and super friendly. Gave me a test to do. Met broader team after VP
Interview questions [1]
Question 1
You’re working with a luxury car brand that wants to optimize their digital advertising spend across multiple channels (Google Ads, Facebook, YouTube, programmatic display). They’ve been running campaigns for 6 months and have the following data: impressions, clicks, conversions, cost data by channel, customer lifetime value, and post-purchase survey data indicating how customers first heard about the brand.
However, there’s a complication: their attribution window shows that customers typically interact with 4-5 different touchpoints before converting, and there’s a significant lag between initial awareness and purchase (average 45 days). The CMO believes their current last-click attribution model is severely under-crediting upper-funnel channels like YouTube and over-crediting Google Search.
Design an end-to-end solution to build a data-driven attribution model that accurately measures each channel’s contribution to conversions. Walk me through your methodology, what statistical approaches you’d use, how you’d validate your model, and how you’d present actionable insights to stakeholders who need to make budget allocation decision.
