AI has moved from boardroom ambition to competitive necessity. Enterprises that can deploy production-grade AI on their most sensitive data, customer records, financial transactions, and patient histories will outpace those that cannot. Yet the majority of AI initiatives stall not because of a lack of talent or technology, but because of one intractable problem: how do you feed sensitive data into AI systems without violating privacy regulations, triggering compliance reviews, or opening the door to data breaches? The answer, until now, has been: you don't. You either expose data and accept the risk, or you exclude it and accept the limitations. Neither option is acceptable for a business serious about AI. Three Roadblocks Killing AI Initiatives Across industries, the same three barriers appear again and again when enterprises try to operationalize AI on sensitive data: 1. The AI-Ready Data Gap Legacy systems weren’t built for AI-scale speed and volume. You’re forced to choose: expose sensitive data (risk) or exclude it (limit innovation). 2. The Compliance Review Bottleneck Without built-in data security, teams can’t verify what AI sees or outputs. Result: endless review cycles and stalled projects. 3. Compounding Regulatory Exposure GDPR, HIPAA, DORA, and a growing wave of new AI-specific regulations demand strict controls over how sensitive data flows into AI pipelines—and auditable evidence that those controls are working. Each new AI workload added without addressing this compounds the organization's regulatory exposure. Doing nothing is not a safe option. Neither is moving fast without guardrails. One Integrated Solution Zero Compromise Cloudera and Bluemetrix have developed a joint solution that removes all three roadblocks simultaneously, without requiring enterprises to rebuild their security infrastructure from scratch. The combination of Cloudera's governed data fabric and Bluemetrix SecureToken, a Cloudera-native Vaultless Tokenization engine, creates a unified security and AI layer that lets organizations do something they couldn't before: feed sensitive data into AI models without ever exposing it in raw form. Because SecureToken is embedded entirely within the Cloudera Platform as a fully certified ISV solution, sensitive data never leaves the platform to connect to an external proxy. All policy governance and encryption key management happen within the same infrastructure Cloudera customers already operate. How It Works: A Real-World Example Consider a data scientist at a European bank tasked with building a credit risk model using customer transaction histories, account balances, and demographic data. Two compliance requirements must be met: The AI model needs the analytical signal in the data—transaction patterns, behavioural indicators—but must never see raw PII, account numbers, or payment records. The organization must provide its regulator with an auditable record proving that sensitive customer data was never exposed to the model in identifiable form. With Cloudera and Bluemetrix, here's what happens: Here is a simplified look at how that request would be fulfilled within an AI environment protected by Cloudera and SecureToken Vaultless Tokenization: Use Cases That Move Forward—Finally Below are the enterprise AI scenarios most commonly blocked by security and compliance concerns, and how Cloudera and Bluemetrix resolve each one:   ... View more

Overview - ExecuteProcess   When developing custom logic as part of our dataflow, having the ability to execute a custom script is a boom especially when it comes to compensating for certain capabilities which are not present out of the box or may not be the need of the hour now for the larger audience. :package:   NiFi makes this possible by providing a processor called ExecuteProcess which enables you to execute custom code. It leverages the underlying operating system to execute the command/process.   Background Over the course of using this processor I often had trouble with managing different modules that were required by different ExecuteProcess processors within my flow. At times I was looking at having different versions of the same modules for different set of scripts. Virtual environments in python came to my rescue here. Hence as a best practise it is always advised to use virtual environments to setup additional modules while using Python code within the ExecuteProcess processor.   Let me now show you how can this be achieved. For the subsequent steps, these are the different versions I am using:   Python Version : 3.6.8 pip Version : 21.3.1 NiFi Version : 1.18.0.2.2.6.0-260 Python Module : Faker   Process The whole process is divided into two stages: Changes on the system where NiFi is installed(Creation of the virtual env) Configuring the ExecuteProcess processor  Stage 1 : Creation of the virtual env         Step 1 : Execute the following command to create a virtual env                   virtualenv <env-name-without-brackets>                   Eg: virtualenv env           *On executing this a directory with the environment name gets created(the name in our case is env) in the current directory                            Step 2 : Activate the Environment and install the package using the following commands                     source env/bin/activate                Note : Once the virtual environment is activated you should see you environment name at the start of the command prompt.           Step 3 : Install the module using the following command                            pip install faker           Step 4 : Once the installation completes, create a shell script(test.sh) with the following content                                  #!/bin/bash                                   source /path/to/your-environment/env/bin/activate                                   python /path/to/your/script/test.py             My script folder looks something like this:                       Important : Make sure the script and the virtual environment folder are accessible by the user which is running NiFi.   Stage 2 : Configuring the ExecuteProcess processor             Step 1 : In NiFi, configure the processor in the following way Your processor will now be able to import the module that you have setup within your python virtual environment     ... View more