Antimatter provides a solution to ensure access control permissions are enforced in data stores and client applications, focusing on making data safe for GenAI.

Antimatter

Antimatter makes data safe for use in GenAI, data science, and other cutting edge apps through a radically new form of access control that integrates in just a few lines of code. 

We’re a team of Berkeley CS PhDs backed by NEA, In-Q-Tel, and leaders at Databricks, Okta, Snowflake, and Google Cloud.

General Catalyst

New Enterprise Associates

Union Labs Ventures

It was energizing to present at the RSA Conference Innovation Sandbox Contest last week alongside nine other security startups doing great work. Thanks to RSA, the judges, and our fellow contestants for a great experience! 

Sharing my short presentation of Antimatter's solution, the Data Control Plane, which ensures data access permissions are enforced no matter how or where data is used. It’s a breakthrough, universal solution for GenAI, data science, and other cutting edge data apps.

I’m eager to hear any thoughts or questions! I’d also be happy to book a private demo – just let me know.

I’m proud to share that Antimatter has been selected as a Top 10 Finalist for the RSA Conference 2024 Innovation Sandbox contest!

The RSAC Innovation Sandbox contest is the most prestigious award for young companies in the security community. This recognition from RSA highlights the importance of our solution, which builds on breakthroughs in AI and cryptography to make unstructured data safe for use in AI and machine learning. 

We’re thrilled to be finalists and lucky to be backed by an all-star group: NEA, In-Q-Tel, and the founders and CISOs of Databricks, Okta, Snowflake, and Google Cloud.

We’d love to meet up with you at RSA. Stop by our kiosk in the Innovation Sandbox room or DM me to set up a meeting.

Mario Duarte Camille Gaspard Andy Konwinski Dug Song Phil Venables John Fiedler Nate Williams Greg Papadopoulos Pete Sonsini Chris Kim William Morrison

Researchers at Northwestern tested 200+ custom GPTs using adversarial prompts. They succeeded in leaking files from 100% of these models, often using basic prompts that require little expertise.

Antimatter’s new data privacy solution addresses these vulnerabilities by automatically classifying, redacting, and tokenizing sensitive data, and by making it easy to apply fine-grained access control policies that apply even as data is moved from place to place.

If AI tools are on your roadmap, DM me for a demo!
#dataprivacy #LLMs

I’m excited to share Antimatter’s revamped website! This update highlights our new solution: a one-stop platform for protecting all sensitive data, no matter where it’s stored or how it’s put to use (data science, machine learning, LLM training, and more). Check out the new demos to see these use cases in action.

If you’d like to learn more or have any questions, please be in touch!

We're excited to share our new AI data privacy solution. If your company is training LLMs with sensitive data, let's talk!

We’re launching a new AI data privacy solution at Antimatter next quarter and I’m excited to share the first demo!

We’ve heard from customers that users are opting out of valuable AI tools due to privacy concerns. We wanted to develop a solution that would serve both parties: strong data privacy guarantees for users & access to redacted and tokenized data for teams building AI.

Our engineering team has developed a really novel solution that will make life easier for AI teams, data scientists, and security engineers. More details coming soon!

#llm #ai #pii #dataprivacy #rag #demo

TL;DR: A sneak peek of our AI data privacy solution 👀 

We hear the same problem from lots of SaaS companies: they build new AI features, but customers express data privacy concerns or even opt out.

We're building a solution that addresses the root of the problem by selectively using customer data for AI while offering strong data privacy guarantees. We have a very novel solution here that I’m super excited about!

See the demo preview below and sign up to learn more. Can’t wait to share more soon!

#dataprivacy #ai #machinelearning #pii #genai

“We’ve had Fortune 50 customers who said they’ve never had a vendor that met their requirements so completely right out of the gate.”
–John Fiedler, CISO and Head of Platform Engineering at Ironclad 

We're proud to share our first case study! Click through to learn how working with Antimatter helped Ironclad become a security leader in contract lifecycle management.

#ironclad #antimatter #datasecurity #byok #encryption #casestudy

We tested Microsoft Presidio and Google DLP, leading ML tools for redacting personally identifiable information (PII). The results were surprisingly bad! Our in-house LLM outperformed both. 

Have you used either tool? What do you want from a data redaction tool? 

#LLM #machinelearning #dataredaction #dataprivacy #presidio #dlp

IO:ANTIMATTER

Antimatter