DataRobot

We use DataRobot for traditional ML/AI use cases from R&D and training to deploying and monitoring. Use cases include churn / acquisition modelling and other propensity models.

We use automodelling (typically OTV binary regression), EDA tools and deployments.

We use DataRobot to generate Machine Learning models that inform the user experience on our website. The predictions are real-time. I have been most impressed with the simplicity in deploying models.

I used Data Robot to design a machine learning algorithm that profiled employee's work environment and absenteeism behaviour (as well as other market factors) to determine if they matched the profile of those employees who had left before us. We were able to use this information to understand the emerging turnover risk of our employees across our various facilities, managers, states and tenures.<br><br>The result allowed us to target our HR initiatives, provide additional training and support to staff and managers, and implement out of the box solutions to newly discovered issues resulting in turnover. It also allowed us to confirm and quantify the impact of different drivers on turnover, which in turn let us prioritise our responses. Finally, we were able to use the models to estimate the impact on turnover and costs a change initiative might cause by looking at the historical impact of initiatives run by our individual sites and/or how difference between a variable had impacted turnover previously.<br><br>Having access to data scientists and project management staff to help design, understand, train and utilize, identify use cases and design the change process was the highlight of their service.

I use DataRobot for predictive modeling. I use this for forecasting and use its recommendation system. I also use DataRobot for automation, to know relevant parameters from the data, and to improve decision making. DataRobot is very useful, it improves our efficiency and productivity because of its automation process such as data processing and data engineering.

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.

This is used for building data models for scorecard monitoring in our department, Risk Management.

I am using datarobot to develop Application and Behavioural Credit Scorecards for the Bank. Develop credit risk models to be used for various business operations (i.e., Products, Credit, and Collections), such as cross-selling, credit limit selling, and collection strategy formulation. Develop credit risk models to elevate lending decision-making and enhance risk management at CIMB PH.

Helping business partners optimize their marketing campaigns based on the strategic objectives of the specific campaigns.

I normally use it for preliminary model checking in order to get an idea to estimate the accuracy or error. This step normally consumes a significant amount of time; nonetheless, it can be determined whether to proceed to the next step of the project or not. My business use cases involve process optimization, anomaly detection, and mainly time series prediction.

We use it to accelerate the building and deployment of our ML platform in our context to solve problems for financial applications.

The DataRobot has great cloud computing and multitasking power for feeding large quantity of data. There are time series, categorical, and numerical related target variable for analysis. The whole bucket preprocessing sometimes could help you to save a lot of times. However in the real scenario it would still be better to do most of the preprocessing manually to avoid confusion regarding data type. The post parameter tuning could be more intuitive and would be better to make the replication of model result easier.

We use DataRobot for a variety of use cases, such as automated quoting, forecasting/inventory management, and automation of machinery settings to improve manufacturing scrap rates.

DataRobot builds machine-learning models to predict creditworthiness of bond issuers.

DataRobot helps us make sense of a large amount of information. Trying to predict what's going to happen is always difficult, but with DataRobot we're able to use the power of robots to do the heavy lifting for us. Let's face it — even if we could do all the complex math on our own, the time it would take would be completely prohibitive. DataRobot takes all that on, letting us get to the decision making instead of setting up models.

DataRobot is part of our DataScience LifeCycle supporting our team in every stage of the AML process. For each business area in my company, data robot contribute reducing the time for development and also reducing the learning curve when we hire a new member in our team. In terms of scope, i have used DataRobot only for batch predictions and it works very well.

Predict churn. Identify ideal customers and targeted customers for promotions.

I use DataRobot to forecast our sales each month. DataRobot builds models from the data we give it and then gives us a prediction on what sales are going to be like for the next 90 days. I then use that data to determine how many products we need to purchase for a particular item (in our case, medical apparel). The balance we are trying to find is having enough inventory to sell so we are not out of stock on a particular item while not having too much of our capital tied up into inventory that is not selling. One problem with selling medical apparel is sorting thru all the data and figuring out what data is consequential or what is not (for example, the size of a piece of clothing is consequential, like Large will sell more than X-Small, but fabric type may be less consequential). DataRobot allows me to use machine learning technology to go thru many different data points and to see what is consequential and what is not.

We use the tool for student success. We make various predictive models for predicting the graduation and retention of our students. We use the predictions to identify at-risk students and refer them to advisors or success coaches for help. For example, if students have less than a 60% probability of graduating within four years based on the predictive model, they will get referred to an advisor for further help. During our use of the tool, we have almost doubled our 4-year graduation rate.

We build predictive models with the core supervised learning product. These include attribution models, churn/retention models, segmentation models, and others. Basically, anything that can be accomplished by taking a supervised, labeled set of training data and turning it into a predictive model, we use DataRobot. We have also dabbled with unsupervised learning and time series modeling but have not purchased those packages.

I use DataRobot to streamline, automate and deploy credit scoring and risk assessment evaluation models for a multinational Fintech with operations across different LATAM markets.

Exploratory data analysis including time series as well as forecasting and building prediction models.

DataRobot is a very useful tool to quickly look for interactions and relations between different features. It serves as a comfortable way to engineer features and rapidly interpret behaviors in between specific attributes.

We use DataRobot as scoring to underwrite new policies. It is an additional help to know who will be a good insurer and who will not. You can use DataRobot to predict something that you need. Our business problem is that it is sometimes difficult to know who is going to have a claim and who doesn't.

We use DataRobot to build data products such as predictive models that are helping the business to move the needle. We use them mainly to adapt consumer experience now that PMI has become a consumer-facing company (it was not the case before IQOS) This year we are starting a new way of engaging with the business to make them part of the full process.