Professional Documents
Culture Documents
Netcore Customer Case Study English
Netcore Customer Case Study English
Netcore Solutions
Netcore’s AI-Engine Raman, built on Vertica, empowers the
marketer with real-time insights and helps maximize marketing ROI
- powered by blazing-fast analytics and highly accurate AI models.
Raman
Contact us at:
Karthigai Muthu www.vertica.com
Data Architect
Netcore Solutions Like what you read? Share it.
“The performance boost from Vertica has Vertica further boosts performance by au- suggestions on how to optimize the content,
been exceptional, with many jobs running tomatically parallelizing the script execution timing, and strategy of campaigns.
more than ten-times faster,” says Debapriya across multiple nodes. The team also uses
Raman provides an entire suite of AI appli-
Das. “And because we no longer need to pre- Vertica’s built-in analytics functions wherever
cations offering solutions around Predictive,
aggregate the data, we can build deeper and required, thus saving additional coding time.
Diagnostic, and Prescriptive analytics. It moni-
more sophisticated models that empower our These features make RAMAN modules more
tors data 24x7 to identify if something is go-
clients to reflect the real world more closely. time-efficient, which was not possible with
ing wrong or if a marketing KPI is taking a hit,
One of our latest innovations—Predictive Netcore’s legacy technology stack.
without marketers having to worry about it, and
Segments—is a very complex feature, which
“Vertica UDx optimizes the AWS data trans- even identifies the probable reasons.
looks at each user’s past web and app activi-
fer time and costs, as well as being easier
ties and predicts their propensity to purchase For instance: Raman is fully capable of identify-
and more efficient for our developers,” says
a product, to uninstall the app, to engage with ing if weekly revenues have started dropping
Karthigai Muthu. “We can also use the machine
a particular campaign, and so on. Vertica’s at the earliest possible stage, and or if revenue
learning libraries provided by Vertica; for ex-
flexibility, built-in analytics modules, and per- from an app has started to fall because of high
ample, we can predict using TensorFlow and
formance helped us build a very time- and uninstall rates. Once the marketer validates
PyTorch models on data residing in Vertica for
cost-efficient solution for our clients. We are and is satisfied with Raman’s observation,
predictive analytics ”
planning to move some of our previous ma- they can use Churn Management to under-
chine-learning work like Preferred Channel For model training, Netcore previously had stand which customers are likely to uninstall
and other Send Time Optimization modules to convert table columns into multidimen- and then target them using a combination of
to Vertica as well, seeing the performance we sional arrays in system memory—a slow and Send Time Optimization, Preferred Channel,
can get out of it.” resource-intensive process. With Vertica’s and Predictive Segments modules to halt the
Array data type, Netcore can directly create trend of decreasing revenue. In short, Raman
Time-Saving Features multidimensional arrays, saving time and sys- is always learning…
Machine learning requires ongoing data tem resources.
transformation, model training, and validation, Internally at Netcore, adopting the Vertica
which creates hundreds of temporary tables Faster, More Intelligent Marketing Analytics Platform has reduced AWS costs,
and large volumes of intermediate data. Using Not only did migrating to Vertica increase increased efficiency, and accelerated the cre-
Vertica’s External Tables feature, Netcore can query performance, but it also cut the time ation and training of new machine-learning
offload these temporary data sets, which required to refresh the various AI models. models. Thanks to Vertica, the company can
keeps licensing costs low. Karthigai Muthu Different Raman modules for multiple clients now bring new functions to market faster,
says, “Without External Tables, we would need can now be retrained or refreshed within a very which helps its clients maintain their competi-
another 6 TB of capacity in Vertica. The decen- short time, paving the way for deeper analysis. tive edge.
tralized architecture also helps us to minimize Karthigai Muthu says, “Our Churn Management Kalpit Jain, CEO of Netcore Solutions, says,
AWS system resources, further optimizing our model has become really fast since we mi- “Netcore has always enabled clients to
recurring costs.” grated it to Vertica. For clients, where end- achieve a better return on investment for
Using Vertica User Defined Extensions(UDx) to-end data preparation and model building their marketing budgets. With our new AI-
has helped the Netcore Machine Learning - used to take nearly a day, we can now com- powered offerings within Raman, helped by
team to run data analysis and processing plete that work within 10 to 15 minutes. This Vertica technology, we are making market-
scripts without the need to pull data out of is amazing as it saves us a lot of time, our ing automation faster, smarter, and more so-
Vertica first. Scripts written in Python can team has already started working on model phisticated. Ultimately, this helps our clients
be registered as libraries in Vertica Analytics enhancements leading to better predictions.” to deliver relevant offers and promotions to
Platform. their customers, boosting both profitability
Netcore clients who combine Smartech with
and customer satisfaction.”
Debapriya Das says, “Since we migrated the Raman not only get faster segmentation—
low-level data analysis scripts to the same in- helping them identify the right audiences for
stances where Vertica is running, this coloca- promoting their brands ahead of their com-
tion made the overall process faster.” petitors—but also AI-driven automation and
03-07-2021 | V | DS | 03/07 | © 2021 Micro Focus or one of its affiliates. Micro Focus and the Micro Focus logo, among others,
are trademarks or registered trademarks of Micro Focus or its subsidiaries or affiliated companies in the United Kingdom, United States and other
countries. All other marks are the property of their respective owners.