Data Science Superpowers

Start using machine learning without the human learning

We believe every organisation should benefit from the power of machine learning without having to invest in a team of data scientists and expert software engineers.

Farrago automates large parts of the data scientist role, empowering non-technical users to apply machine learning algorithms to existing data with a few clicks of a button.

PREDICTIONS IN 7 CLICKS NOT 7 WEEKS.

NO CODING. NO PhD REQUIRED

Power Up with Farrago START DEMO

what you want

  • Quick answers to questions the business needs for decision making
  • To start on the journey of utilising AI and ML without massive investment upfront
  • To empower your analysts and BI teams with data science capability

what you don’t want

  • To spend $000’ks on building a team without knowing when it will pay back
  • Wait for weeks to get predictions back that are needed in days
  • A huge programme of work to get cutting edge technology into the business

Your journey to

data science superpower

STEP 1
STEP 1: DATA & QUESTIONS
Import your data and Farrago immediately gets to work. Using proprietary natural language algorithms and statistical analysis, Farrago sufficiently understands your data to filter possible prediction use cases and guide you on what columns could be useful to consider.

Data scientist time: 1 week

STEP 2
STEP 2: ANALYSIS
Driven by your use case, Farrago analyses your data to present what needs to be done to make it usable in the machine learning model that matches what you want to predict. Each model requires data to be wrangled in a very specific way. You receive a report on exactly what needs to be done.

Data scientist time: 2 weeks

STEP 3
STEP 3: FIX & TRANSFORM
One click and the equivalent of a team of data scientists and data engineers get to work on your data, fixing and transforming it for machine learning. Then a whole load of code gets to work to ready your data for training, testing and validating your predictive model.

Data scientist time: 2 weeks

STEP 4
STEP 4: RUN MODEL
Machine learning models require a whole heap of computing power to crunch all the numbers. Farrago has custom pipelines to integrate directly with AWS Sagemaker for training, testing and validating your predictive model.

Data scientist time: 1 week

STEP 5
STEP 5: RESULTS
One of the big challenges with machine learning is interpreting the results, which typically are entirely meaningless to even above average humans. Farrago converts the model outputs into a report that can easily understood by analysts and business stakeholders.

Data scientist time: 1 week

Predictions

Here are some use cases and data sets that are available in the Farrago demo. The data quality is realistic; it has missing values, redundant columns and bits of information that would usually make it hard to run a prediction using machine learning.