An ideal end-to-end AutoML solution and force multiplier for skeleton crews!
October 07, 2023
An ideal end-to-end AutoML solution and force multiplier for skeleton crews!
Score 9 out of 10
Vetted Review
Verified User
Overall Satisfaction with DataRobot
As Australia's largest radiology business, we use it as part of our sales and outreach program to identify referring doctors who we need to prioritise contact with for a variety of goals including retention and growth. The models trained and deployed on DataRobot assist with this prioritisation process. The output of the model is effectively a risk or opportunity score on how likely a doctor will either increase or decrease in referrals to our clinics if we do not check in on them.
- autopilot for testing and ranking the suitability of multiple models
- easy to upload observations and download the predictions with explanations
- scalable with multiple workers to speed up the process where urgent
- more lenience with uploaded observations when the feature tables don't fully match the feature set that the deployed model trained on. Instead of simply erroring out, provide prompts for in-place fixes.
- null correlation analysis (how often two columns are missing values at the same time) would be very useful to help identify a different type of data relationship.
- Helped with loss prevention; measureable value
- Influenced workloads by our sales team
- Improved impact per contact hour by sales team.
Yes. I would recommend using DataRobot for any new organisation that has a skeleton crew in Data Science and are very early in their technology maturity level. It removes the costs related to engineering and MLOps and allows the scientists to focus on target design and model review rather than configuring for and executing the training and evaluation process which, outside of data cleaning and ETL, is the largest time consumer that should be an automated and reusable pipeline for 95% of scenarios.
Patient data is highly sensitive and limits the use cases we can experiment on and deploy to DataRobot. I'm aware they offer some promises on privacy and security. For example, without more guarantees and clearer compliance measures, we won't be able to use it for clinical applications such as image processing of MRI scans to identify anomalies and augment radiologist reports.
As such these opportunities are unlikely to be worked on until we have policy and framework changes to upload imaging data to a cloud service that has servers outside of Australia.
As such these opportunities are unlikely to be worked on until we have policy and framework changes to upload imaging data to a cloud service that has servers outside of Australia.
- H2O and Amazon SageMaker
Comparable to H2O but my company chose DataRobot so that's why I'm using it. Pricing is reasonable and the feature coverage is probably better from an end-to-end perspective. DataRobot has less flexibility than Amazon SageMaker but is a lot simpler to use, which again for a small team is beneficial and a force multiplier in the short term until the organisation matures and invests in largest in-house IT teams to support the AI programs.
Do you think DataRobot delivers good value for the price?
Yes
Are you happy with DataRobot's feature set?
Yes
Did DataRobot live up to sales and marketing promises?
I wasn't involved with the selection/purchase process
Did implementation of DataRobot go as expected?
Yes
Would you buy DataRobot again?
Yes