Cambridge, Massachusetts headquartered Akkio, which claims to offer a no-code platform to help enterprises build and deploy AI in minutes, today announced notable product improvements including integration with major data platforms.
The performance of an AI model is only as useful as the data it has been trained on. However, enterprises today rely on different sources and applications for different data points. This makes model training a tad difficult. With this update, Akkio is solving this problem, giving organizations a wide range of options to connect to for training models and getting insights, including Snowflake, Google BigQuery, Airtable, Hubspot, Salesforce, and Google Sheets.
“We’re trying to make AI as easy to use in business as Excel spreadsheets. Now anyone can get quick data wins — using AI to surface patterns and optimize key business outcomes,” Jon Reilly, cofounder and COO of Akkio, said.
Features for data preparation, anomaly detection
Beyond offering integration with a swath of data sources, Akkio’s platform is bringing capabilities to help enterprises better prepare their data to get accurate model results and detect anomalies in those models.
For the former, the company said it is adding filtering and merging capabilities, which allows companies to weed out unnecessary data rows and match records without unique IDs. Meanwhile, for the latter, it is offering custom models that can detect anomalies in data allowing preventative maintenance on IoT devices and detection of fraudulent transactions among other things.
Akkio is also introducing time series models, which look at collections of observations in chronological order to surface data patterns over time and forecast likely future outcomes. This usually helps with use cases such as churn reduction, forecasting quarterly revenue numbers, and weekly inventory figures, making any organization’s data more valuable.
Ten times faster insights
In addition to this, the platform is bringing the ability to help analysts see the most predictive factors in their data and understand the combination of factors that could drive outcomes. It automatically creates clusters that group together cohorts of data around an outcome, which can then be used by businesses to target specific cohorts with different offers. This, Akkio claims, can optimize their conversion and drive insights ten times faster than before.
The updates strengthen Akkio’s offering in the no-code AI development space, which has been growing particularly in light of the pandemic and the shortage of data science talent. Other players operating in the same segment are Google AutoML, Obviously AI, and Fritz AI.
Gartner predicts that 65% of app development will be low-code/no-code by 2024.
0 Comments
Enregistrer un commentaire