regardless of the size of the machine you're using, TIMi/Anatella just delivers:

- I've loaded about 1TB of compressed Json documents straight into the tool

- the transformation to an actual matrix of features was by far the quickest I've used (vs R vs local SQL db vs Python)

- ability to multithread or parallelise which significantly reduce the execution time

- ability to build end to end pipelines without the need for a third party tool

- large range of connectors (for input or output) making integration trivial

- very fast to iterate and test new approaches

For our needs, TIMi is just making any other solution irrelevant both from a performance and from an economic perspective (no need for big and expensive clusters with dreadful admin and workload management costs)

The UI might seem slightly confusing at the beginning but it needs to be mitigated with the fact than other "alternative" solutions are exclusively code based

Large scale data engineering and machine learning.

Again, it's the fastest and the most economic solution I've ever used given the size of our datasets

Intuitive, simple tool with processing power not seen in any other tool at the personal pc level. Constant evolution of the processing nodes and native integration with many of the commercial data warehousing platforms.

Kibella could be much more user friendly, it is not as simple as Anatella, Stardust and Timi Modeler.

An online library of examples or scripts with basic functions could be incorporated for the benefit of the community.

Try it, you won't regret it

ETL processes, analytics for energy fraud detection models. Telemetry processing

The large number of actions available to manipulate data. You have options to tranform data without programming, but if you want to program new actions it's possible too.

The applications works only over Windows

I have automated several processes in my work, from automatically reading emails and taking actions, to comparing information from my clients to validate data quality. The tools for text mining are my favorites.

The speed and the user-friendliness. You can easily automate in a few mouse clicks the most complex data processes: data cleaning, computation of kpis, scoring, etc. You practically never have to write one line of code (but you can also code if you really want to)

I already received a data transformation made by a colleague that was leaving the company. This data transformation was complex (more than 100 nodes). So, it was difficult to understand. I regret that Anatella does not impose by itself any boundary of the complexity of a transformation. This might lead to a difficult situation for maintenance of very large transformations. On the other side, the visual representation of the transformation is really helpful to understand what’s going on.

Data automation, Data cleaning tasks, Reporting & Scoring.

- The graphic modules implemented make a great UI, instead of only code based.

- The flexibility to make changes in ETL's already working.

- Really fast processing speed for a big loads of information.

- You don't need a really big server o machine, TiMi simply works everywhere.

- A detailed user manual built in Anatella and the easy way to check errors.

- A large connectors options, for input and output, giving Anatella a wide spectre of usability.

- You can read basically every type and any quantity of inputs.

- We always have a great customer support.

- It's not a well known software, which means you won't be able to find people who knows how to work with Anatella and there is not much forums or communities where you can ask and share your experience.

- The licence cost could be a problem if you don't have a team or business model that extract all the bennefits from the software.

Having clear your needs and your cappabilities to have all the benefits from the software.

- Optimization and automation of processes, reducing the time needed to acomplish repetitive task and reducing the cost implied in this (people, penalizations, customer satisfaction).

Speed in terms of pipeline development in Anatella; creating a complex data pipeline is ten times faster than programming it in Python. Even for experienced programmers.

Speed in terms of model training in TIMi, after the data preparation phase, creating a classic binary prediction model is done is n less than one hour, allowing for multiple iterations.

Speed in terms of data processing. never seen anything this fast especially on large (100 mio plus records) datasets.

Because it is still a small company it is not (yet) taken seriously by IT and procurement departments. So when you work for a larger company or want to use it for production proposes, IT often says no.

Doesn't work on Mac.

Ask for a trial licence and experience it for yourself.

I've used TIMi in many Proof Of Concepts of machine learning applications in Customer Values Management, Marketing, Sales and Distribution. Using it to process large amount of network data, building a training datasets and use TIMi to create multiple models. Benefits are rapid development, to option to fully automate model building, allowing to auto create over 100 models for an NBA application.