LightfieldAI-native CRM that builds itself and does work for you
SecureIntent
Stop credentials from reaching AI. Protect every prompt.
18 followers
Stop credentials from reaching AI. Protect every prompt.
18 followers
Secureintent is due to launch on the 10th of July. What makes our solution unique is its a free chrome extension tha acts like a DLP cybersecurity tool however its zero retention, meaning sensitive credentials are never stored. This aligns really niceley to developers who want a guideline, not a blocker in thier workflows. Our founders all come from enterprise security backgrounds, meaning we take the best industry expertise to create a tool for developers.
How does the zero retention actually work in practice, like does the extension analyze data locally on my machine or is it sending anything to a server for processing?
@denizpalantblu Locally, thats why its fast and lightweight.
Nice angle on staying out of the way. One thing that would help my team is adding a lightweight Slack or webhook alert when the extension blocks something, so devs get a heads up in the channel they already live in instead of only finding out after the fact.
@emircan67n3 Hi Emircan, agreed, yes, team features will be available from next month. Would you like a lifetime Business Pro license to test the functionality in Slack?
the zero-retention angle is the right call for something sitting between devs and their prompts, nobody wants a security tool that becomes its own data risk. curious how the detection itself works though - is it pattern/regex matching against known credential formats (API key shapes, AWS keys, etc.), or entropy-based scanning for anything that looks like a secret regardless of format? asking because the failure modes are pretty different: regex misses novel/custom token formats, entropy-based flags a lot of harmless high-entropy strings (hashes, IDs) as false positives. which side did you optimize for?
@galdayan Hi Gal, great question. what makes you ask?
@secureintent_ai mostly curious because I've seen tools claim entropy-based detection and then choke on things like base64 hashes or UUIDs in logs, which trains users to ignore the warnings. if you're doing pattern matching first and falling back to entropy only when the shape is unknown, that's the combo I'd trust more than either alone
@galdayan Hey Gal, thats actually the way we are delivering the solution. You are the first person to pick this up, clearly you are very intelligent.
@secureintent_ai appreciate the kind words but that didn't quite answer it - is it pattern-matching first with entropy as a fallback, or the other way around? no worries if that's not something you want to detail publicly, just curious for my own understanding.
How does the extension actually scan for sensitive credentials if nothing is stored, does all the pattern matching happen locally in the browser?
@aleynarr7b Check out our landing page secureintent.ai it tells you there and our code is auditable.
The zero-retention angle is genuinely interesting, especially for devs worried about pasting secrets into random fields. Liked how it flags without blocking you mid-workflow.
@etinynylac4u Thanks Cetin
love that you went zero retention from the start instead of bolting it on later, that is exactly the kind of choice that shows the team actually understands enterprise security rather than just selling it.
@dnegayretztem Thankyou, agreed.