Questa releases a Privacy focused AI Analytics Assistant that first anonymizes all sensitive information from documents to prevent AI training on them.
SAN FRANCISCO, CA, UNITED STATES, January 3, 2026 /EINPresswire.com/ — As artificial intelligence embeds into modern daily work—from business analytics to AI Agents running automations in sales, operations, media communication and payments —concerns around data privacy have never been more urgent. The rapid adoption of AI has already led to promising productivity gains—but it has also introduced new risks. Many AI platforms collect user prompts, store conversation histories indefinitely, or use customer data to retrain models without clear consent. For businesses, educators, and individuals alike, this creates serious compliance, confidentiality, and trust challenges.
In an era when AI tools often rely on extensive data collection, opaque model training, and third-party data sharing, users are increasingly asking a simple but critical question: Can they use AI without giving up on their privacy?
Questa Safe AI tries to answer that question with a clear and confident “yes” as they introduce their new and unique Safe AI Assistant called Novara.
Novara empowers users to find answers and insights from their own everyday business files using any AI model like GPT, Claude or Gemini without the risk of AI Training on confidential information. To prevent AI training on sensitive data, Novara first anonymizes all files locally in a Blackbox – either on the enterprise server (called on-premise) or in an exclusive cloud account belonging to the user.
Questa itself provides more AI based services like Business Analytics but the Blackbox Anonymizer works for all 3rd party Agentic AI services too.
Built from the ground up with privacy as a core design principle, the Blackbox replaces all personal and business critical information in all the user documents before allowing any AI models like GPT, Gemini, Claude or any AI Agents to ingest them.
Users can run Questa in an exclusive cloud account that they can set up through the company’s website or install in their enterprise server from where the files are redacted like an automatic firewall before any AI models can touch them. Users cannot accidentally send their data to LLMs and Agents as both of Questa’s models – cloud or on-prem versions are designed to anonymize files before any Generative AI services can access them.
LLMs are already highly commoditized with different ones showing strong aptitude for different tasks depending on their internal training set. Hence Questa was designed in a way where users can choose different LLMs for different analyses to evaluate their relative strengths for a task without locking into one specific service provider.
After the Blackbox anonymization, users can query and find deeply intelligent answers in their own excel, word or pdf files using Novara. Users can also generate reports from unstructured documents in financial analysis, sales, marketing, strategy and customer feedback in minutes instead of excel modelling and other analysis which took days.
Unlike conventional AI tools that prioritize scale and data harvesting, the platform follows a privacy-by-design approach.
Key privacy principles include:
– No API pulls by 3rd party AI Models before local anonymization
– No prompt logging
– No training on user data
– A two layer model to locally remove all sensitive information.
– Enterprise grade security with on-premise installation and national data residency obligations
For interested readers in the core technology, they can read this under the hood research article by Questa Data Science team.
The founders of the project are serial entrepreneurs from ex-NATO, Amazon Web Services, SAP Labs and European health-tech background. The project advisors include Professor John MacIntyre, one of the pioneering AI Scientists from the UK and a leading voice on AI Ethics.
The founding team has deep knowledge in critical financial and healthcare patient data privacy and have built the system with enterprise grade security especially for critical sectors to prevent unconsented usage of sensitive data.
To mark the New Year and make privacy-first AI accessible to more individuals and teams, Questa Safe AI is also offering a limited-time 75% holiday discount on its plans after starting a free trial.
David Cooke
Questa AI
david@questa.solutions
Visit us on social media:
LinkedIn
Instagram
YouTube
X
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
