0:00 Then the next one is the data loss prevention. 0:03 So again you know we have added additional predefined data identifiers like So predefined data identifiers are like Social Security numbers, passports or numbers or or or or or date of birth or anything which is sensitive, OK or, or something relating to medical field, OK. 0:24 And so we have added more predefined data identifiers and then we have also added predefined data profiles like some of the predefined data profiles like PCIDSSGLBAGDPR. 0:36 So we added more for various industries and various geographies. 0:40 And then we have added more activities like tokenized forensic, a legal hold and then support for not operation. 0:48 So when you are defining like if you are defining you can define a DLP rule and say that that if a document has got a date of birth and credit card but it does not have something else then only that rule gets activated. 1:08 So we are at added support for not operation and then we have also added support for AIML for identifying both pictures. 1:19 So here what happens is that like if somebody is making a picture of a driver's license or a circuit board, a circuit board or AutoCAD images. 1:28 So the driver's license is what we call a structured. 1:31 Structured means because there were certain pattern. 1:35 When you look at California licenses or or or Nevada licenses, there were certain pattern. 1:42 But when you look at circuit boards, schematics of buildings or HVAC systems, so these are unstructured. 1:50 So we use AIML for, for, for identifying this, for identifying both unstructured as well as well as structured images and for preventing exfiltration. 2:05 So last year, I mean since the last now 1 1/2 two years, we have released exact data match, which is the useful to many companies. 2:18 And we should, we did show you the screenshot for those. 2:21 They are always available. 2:22 They're already available in our VCG. 2:24 So today I'm just going to show you screenshots relating to OCR as well as fingerprinting. 2:32 So what is the document fingerprinting? 2:35 So document fingerprinting is different than fingerprinting which Michael talked about, which was the device fingerprinting. 2:42 So in document fingerprinting, like let's say that you have insurance form your insurance company and you want to make sure that none of your employees, they exfiltrate any information. 2:54 So what the admin does, he takes a blank insurance form and he and he uploads to the WhatsApp cloud gateway through the portal. 3:04 And so we tokenize everything within that document. 3:09 And so it creates some kind of a fingerprint of that. 3:13 And now if somebody else like person with bad intent, what he wants to exfiltrate information, so he's uploading some document which has which has some sensitive information like maybe Social Security numbers or, or, or or date of birth and all of that or something else. 3:36 Then then again, we do a fingerprint the Warsaw cloud day gate. 3:39 We also does a fingerprint of that and it makes sure that that that the representative document and the one which is being exfiltrated as the same tokens and something more than it basically then it flags that event and you can then decide to block that. 3:58 So that is document fingerprinting. 4:00 So we have added support for that and and then Microsoft, Microsoft Information protection or the Azure information protection where you can have labels to document and based on that we can prevent exfiltration. 4:15 So now we have added support for even something called as protected label. 4:18 So even if a document has protected label we can rewrite them. 4:25 So here again, you know the SoC admin, what he does is he uploads the representative document which he wants to make sure that the empty document is OK to send out. 4:37 But if somebody is got some sensitive information within that to stop it. 4:43 So the user, the admin first of all uploads using the Concerto File Manager the document and then he creates the DLP rule for fingerprinting and a DLP rule for OCR. 4:55 So even if he sends the image of the document, we'll be able to stop it. 4:59 And then he creates the DLP profile with these rules, the fingerprinting and OCR rule. 5:04 Then once he commits it, then what would happen is that we would and then the when the user uploads the sensitive document like here, then again we would go the Warsaw Cloud Gateway would create a fingerprint of that and compare. 5:24 And we have something called as a similarity threshold. 5:27 So the empty boxes and the filled boxes, they will be compared and based on the similarity threshold, we will flag it and stop it or do whatever you want, encrypt it or warranty in the document. 5:42 And the versa will also do OCR if required. 5:47 And so that's what and then compare the fingerprints of the received document and the representative document. 5:55 So here this is like a screenshot of our File Manager. 6:02 So you can see that. 6:04 So whether you want to upload EDM relating data which is the exact data match or fingerprints or address files or certificates or URL match files. 6:14 So you can one can do using this file the files and folders. 6:19 And so in this particular case there is this application form.pdf which is the that's the blank form for which we want to prevent exfiltration. 6:31 So that's what the user has uploaded and and this is what like a DLP rule corresponding to fingerprint looks like. 6:40 So the the user specifies what we call as a similarity threshold and if this is if this is violated then then the block action is taken in this case. 6:54 So this is for what DLP. 6:56 So in short, we have added support for DLP based on AIML fingerprinting, support for protected labels, EDM everything which you already have. 7:08 And we are also using now NER for a named entity recognition for DO, DLP and some other AML models as well.