Frontier, like many of our clients, has to endure the never ending challenge which faces all analysts – managing the data.
A new way to model
Complex models require detailed input data and produce a wealth of output data. This data needs to be verified as up to date and error-free, versioned, stored, referenced, analysed and interpreted. Once the inputs have been collected, they are transferred into any number of models or tools, then (numerous) modelling runs follow – the outputs of which also have to be managed. All of this needs to occur continuously, usually under tight timelines and without any mistakes!
Developing and enforcing the processes and mentality needed to deliver all this is costly in resource terms and, of more concern, sometimes limits how much analysis can be performed to inform critical decisions within a given timeframe.
Our response to this challenge has been to adopt best practice approaches used in web-based application development. This has resulted in Frontier’s Knowledge Integration Technology, or ‘Kit’. Frontier’s ‘Kit’ represents a generalised system for combining a number of data sources, models and analytical routines within a robust IT architecture.
This underlying architecture is presented via a powerful and intuitive, web-based user interface that allows multiple users to upload and modify centralised inputs in a managed way, to build, save and run modelling cases and to gain access to a range of customised reports and raw data outputs. The platform is flexible and scalable allowing for changing requirements, availability of new data and increasing analytical sophistication over time.
‘Kit’ is helping our clients to undertake analysis that once took months in days. Problems that clients are addressing using ‘Kit’ include:
- determine a combined power station maintenance and trading strategy for the next year
- conduct long term modelling of market outcomes to inform policy reviews.
As part of this process, we have been able to use the core economic frameworks from our existing models and apply them to new problems. This has lead to the development of our new models ‘Whirlygas’ and ‘Splash‘.