Ople Simplifies AI for the Data Scientist


Ople, the San Mateo startup is on a mission to provide AI solutions that are “easy, cheap, and ubiquitous” to data scientists around the world. Pedro Alves, who is working on his Ph.D. in computational biology founded the startup in 2017. He believes that building, testing, and deploying AI shouldn’t take months but days. And this belief has guided him and his team in doing just that.

The startup focuses on infrastructure plumbing, simplifying laborious tasks such as data preparation, feature engineering, algorithm selection, testing, training, and so on. Ople has the entire AI process down to a science. First, the customer prepares the data, uploads it, then Ople takes over from there in optimizing, building, and deploying the model.

Ople AI Process

  1. Preparation: Convert dataset into CSV or XLS
  2. Upload Data: Upload data files
  3. Validate Data: Check column format to make sure it’s ok
  4. Configure Optimization: “Name model and set max running time”
  5. Model Building: Platform applies custom feature engineering techniques and algorithms on the dataset, and “generate a custom final AI model”
  6. Deploy: Model “starts making predictions”
  7. The Blackbox: Visualizations offer deep insight into the results

Although AI is the growth engine that will drive the digital business for the foreseeable future, it has its limitations. During a panel discussion, Tom Annau, Senior Director, AI, and Advanced Architectures at Microsoft discussed how an algorithm mistook a Stop sign for a 45mph speed limit sign. Apparently, there was graffiti on the sign and the model decided to remove it from the sign, thus causing confusion for the model. Because of incidents like this, Ople is focused on being transparent about how models are making predictions.

In another case, the UCI Department of Computer Science evaluated the results of a black box machine learning prediction on whether a dog was a husky or wolf. A student developed a classification model, and it was able to predict the precise animal every single time. However, taking a closer look, the model was making the decision based on the background, in this case, snow or no snow, and not the animal itself. Sameer Singh, Assistant Professor at UCI states that machine learning is everywhere – email, phones, games, etc., thus, how can it be trusted. Transparency is important when dealing with AI and every organization starting from Google on down has a role to play in making sure decisions are being made in the right way.


  • Company: Ople Inc. 
  • Founded: 2017
  • HQ: San Mateo
  • Raised: $10M
  • # of Employees: 57
  • Founder: Pedro Alves (CEO)
  • Product: AI platform for data scientist that simplies the process of working with AI
Scroll to Top