Factored built a user-friendly pipeline so that images from Google Street View could be utilized to predict the feasibility of building aerial fiber on specific roads. This delivered insights to a Private Equity Fund that informed the acquisition of telecommunications (telco) assets for $7.5 billion USD.
The Challenge:
To retrieve images from the internet given a set of coordinates for a telecom company to assess the feasibility of building aerial fiber in new areas.
The Solution:
Factored built a process of checking image availability and estimating the costs associated with downloading the images. We produced CSV files containing the predictions of the feasibility of building aerial fiber on specific roads. The company was able to use this information and apply it to advise on the acquisition of telco assets for $7.5 billion USD.
The Outcome:
Et sapiente blanditiis 33 sunt accusantium et quasi voluptas. Et vero possimus qui dolorem beatae vel blanditiis galisum non adipisci animi. Ab suscipit Quis et earum dolor et sint ipsam 33 nobis quidem est laboriosam 33 sequi voluptatem.
Tech Stack & Skills:
PySpark, Google Street View, PyTorch, fast.ai, Boto3.
Parallel Computing, API Integration, Image Classification, Model Transformation, Module Packaging.
Results
Stuff
Stuff
Stuff goes here