ZeroTrusted.ai

Shield Your AI Privacy & Security with Zero Trust

471 followers

Shield Your AI Privacy & Security with Zero Trust

471 followers

Visit website

Safety and Privacy platforms

•

Security software

It offers robust LLM Firewall solutions, ensuring your data and identity are protected in the AI space. Our service secures your data, offers prompt anonymity, and prevents unauthorized access. Perfect for users valuing security.

Free

Launch tags:SaaS•Privacy•Artificial Intelligence

Launch Team

Framer AI AgentsDesign and publish professional sites with AI

Promoted

ZeroTrusted addresses 2 major pain points of the LLM, 1. Privacy 2. Ensembling of multiple LLMs Amazing, looking forward to dive deep into the ZeroTrusted.

Report

2yr ago

ZeroTrusted.ai

Maker

@siddharth2395 Yes, Privacy and LLMs summarization are the main points. Thanks for the review!

Report

2yr ago

Jessica by Queue

As someone with clients in law, finance, etc. (highly regulated industries) who value privacy and data security, we've been unable to fully embrace AI technology like ChatGPT because, in line with their internal policies, we cannot input any of their company data into such LLMs. I like that this platform allows us to harness the power of language models without any concerns about data being tracked or misused. It's a game-changer for teams safely and easily using AI applications. Great idea! I wish I'd thought of it! @sidraref @vijay_karavadra I like that in your chat you have some examples that I can click on to see the product in action. A great launch, team!

Report

2yr ago

ZeroTrusted.ai

Maker

Hey @jake_mcmahon1 , I'm glad you liked the idea and it's useful. Give it a try and drop an email if you need any help and we would be happy to help - contact@zerotrusted.ai. Thanks.

Report

2yr ago

This web app seems to be promising to provide features for data security. What are measures employed for data security to ensure compliance with data protection regulations and safeguard against potential breaches

Report

2yr ago

ZeroTrusted.ai

Maker

@bilal_mohammed1 Good question. We detect any PII/PHI/PCI in your provided search prompt and sanitize them before passing it over the LLM. We also prevent a list of compliances like NIST, HIPAA etc. At the end of chat, you get a detailed report which shows detected and fixed privacy issues.

Report

2yr ago

We provide software development and mobile app development please if you need any IT solutions services ping me on jaydeep@vnnovate.com

Report

2yr ago

Opengeni

Are you using OpenAI ? or open source models or?

Report

2yr ago

ZeroTrusted.ai

Maker

@sentry_co we support openAI and many other LLMs. In our premium plan, we enhance accuracy by running your query on multiple LLMs and then we use our scoring model to determine most reliable - think Kayak. Users can also select 1 LLM. Our main role is to be the privacy layer for our customers.

Report

2yr ago

Opengeni

@femitfash If data winds up at OpenAI. Then privacy is broken. There is only so much you can do with obfuscation. Or am I missing something?

Report

2yr ago

ZeroTrusted.ai

Maker

@sentry_co this is partly true, however: 1) any sensitive information are already masked by our system. 2) most importantly, your identity isn't tied to that query you searched and it cant be traced back to you since we don't keep any data and wont even know who sent what. Your queries are just so anonymous text embedded within large corpus amount of data. BTW - this is only aspect of the value we provide.

Report

2yr ago

Opengeni

@femitfash Right. But human data is extremely identifiable. And if you obfuscate it too much, then OpenAI doesn't understand it right? or? Im just trying to understand here. Some kind of illustration or example would make this more clear I think. A segment of text goes into your service then goes to openai etc. What does the data look like at each juncture etc? You know what I mean? Btw. Love the problem your trying to solve here!

Report

2yr ago

ZeroTrusted.ai

Maker

@sentry_co Here is an example - suppose a medical expert needs to process the following input: "Create a summary of this patient's diagnosis: Patient Name: Paul Smith Date of Birth: 12/08/1987 SSN: 666-555-5555 Diagnosis: Chronic Heart Disease Treatment Plan: Undergo heart surgery by Dr. Smith at General Hospital on 03/15/2024 Insurance: ABC Health Insurance, Policy Number 434444444" Before we pass it on to any LLM, we'll alter the sensitive (PHI compliance violations) details and transform to: "Create a summary of this patient's diagnosis: Patient Name: John Doe Date of Birth: 01/01/1970 SSN: 123-45-6789 Diagnosis: Chronic Heart Disease Treatment Plan: Undergo heart surgery by Dr. Smith at General Hospital on 03/15/2024 Insurance: ABC Health Insurance, Policy Number 987654321" This way, we preserve the context, while ensuring that your PHI compliance is met and sensitive data isn't exposed to LLM. After getting a response from the LLM, we revert the swapped sensitive value to your original while preserving the result context. We will be adding features that will allow RLHF to learn and adjust to customer preferences. Hope this helps.

Report

2yr ago

What criteria do you use to determine the reliability of LLM outputs?

Report

2yr ago

ZeroTrusted.ai

Maker

@asad_mobile We use below criteria, 1. Hallucination 2. Plagiarism/Copyright violation in LLM responses 3. Prompt detection % of each in the calculation are adjustable to customer needs.

Report

2yr ago

Congrats! Very cool solution! How are API keys being handled by ZeroTrusted?

Report

2yr ago

ZeroTrusted.ai

Maker

@charlestehio we have an identity server that generates the keys. 1) Only the hash of the keys are stored in our system so that bad actors can never get hold of it. 2) When a user makes an API request the key they provided will be hashed first and then compared against the saved hash for authorization. 3) If the key is lost, a user will have to regenerate a new one. 3) the keys are bearer token so you have to ensure you keep it safe don't accidentally expose it. With regards to our end to end encryption, I can not go into too much details as some of our approach is part of our intellectual property :). But we provide some additional overviews on our site. you can check it out. We also have layers of protection to monitor behavior pattern so as to detect if the request is being made this is available in our enterprise plan. Please message me directly if you have further inquiries.

Report

2yr ago

•••

3 4 5

•••

Reviews