The way Jiffy.ai integrates into existing infrastructure has been great for us. Our company is pretty stringent when it comes to cyber security and integrating with our apps. For every automation that we do, we have our technical architects involved especially since, when we implemented this a few years ago, it was really something of a new technology, knowing that you've got "robots" accessing systems and updating records and altering information. It's a little bit daunting if you're not familiar with it. We've definitely had very strong scrutiny over this platform and this work, and even within that, it's been really successful at being able to integrate.

You still have to integrate with, or at least access, the systems that you're automating within. For example, if you're doing something within SAP, you're still going to need to access data or screens or APIs or something to interact with that system. But the fact that the solution incorporates intelligent document processing, among other features, means we don't have to integrate with another document processing capability or technology. That's a big reason we chose Jiffy.

The most important part is how easy it is to pair the automation. So, it is a canvas that is just drag and drop. You don't have to code, so it is a no-code to low-code solution.

It is good for simple tasks that we have done in the past, e.g., reading the invoices. A valuable feature is the document processing. Usually when we talk about document processing in the market, you just have OCR. Where once you extract the information, you need to program or do some type of data wrangling to actually get the value of it or process it. For Jiffy.ai, they have the machine learning behind it, so we didn't need to code one by one. For example, if you have 5,000 vendors who are sending you different types of invoices, then we are not talking about 5,000 invoices. We are talking about one vendor who has three types of templates, so that is about 15,000 documents to process. Even if you do OCR, you want to extract the information and code it to read this and that. So, Jiffy.ai has machine learning where we don't have to teach all the documents, instead we just need to teach it a few. Then, the machine will already know if it finds this type of information, then that is what it is. For example, the easiest way is the invoice number. Most vendors usually have similar wording: invoice number, invoice NO, and INV. However, in all 15,000 documents, you see that the vendors just play around with this wording. It won't differ much. Therefore, the machine learning knows because of this, you don't need to teach it all 15,000 documents. After about 10 documents, the bot can pick it up themselves and learn about it.

There are not a lot of vendors in the market who provide built-in machine learning. In the invoice, you have multiple things that you want to extract: invoice number, PO, and some other line items. With machine learning, we expect it to know what to extract from, by looking at different templates of invoices. It should know that this is similar. Even though you use the different wording across multiple templates, the machine should know that it is an invoice number. We expect the machine learning should be able to do this, and the Jiffy.ai machine learning is able to do it with 80 to 90 percent accuracy. So far, we haven't had a big problem in whatever the machine learning reads, doing it correctly. If it didn't read correctly, we would have to correct it, then the bot will learn from that, "Okay, this is actually the better way," so it can do better next time.

The workflow engine is definitely a very strong asset of Jiffy.ai, because it is easy to configure. It has a nice user interface. It is also scriptable. It doesn't have a steep learning curve. It is quite easy to learn, so you can become productive very quickly. Up until now, their automation tool combined with the workflow engine has been their strongest asset. It has helped us extract information out of an application which otherwise would have to be done manually. So, it gives us the opportunity to automate a lot of tasks related to extracting information, rather than delegating that to actual people. It has saved us hundreds of hours per month. It has covered the work of two or three full-time operators.

Jiffy.ai's app-based approach is suitable to automating entire complex business processes and to an approach that only automates specific tasks within a process or workflow. My impression is that the solution is so flexible that it can combine multiple applications into the workflow and interact with all of them. For example, in a Windows environment, it can launch one UI application, interact with it through the workflow, launch a process into a remote virtual machine (or interact with a remote service), fetch the result, and then feed this back into the local desktop application. My understanding is that it can deal equally well with tasks within a single process and tasks that span multiple applications in multiple environments.

It can definitely support integration with other third-parties. The combination of all these features can create very powerful applications. Our use of Jiffy.ai so far is a bit limited because we are using desktop workflows and the AI aspects of it. However, combining these can create a powerful set of features for creating more advanced applications. It can be an integral part of a bigger system. For example, you can have a front-end application that is delegating requests back into the Jiffy.ai, then Jiffy.ai will essentially act as the orchestrator for back-end services. So, it's quite powerful. The fact that it has a UI means it is accessible to non-technical people as well. So, you can get from the design phase to implementation phase very quickly.

Jiffy.ai has its own notations for specifying the theme navigation of individual nodes. That notation has the most common structures that you would expect from a programming language without some of the most complex features, like memory management or complex design. I feel it is accessible to junior developers. Now, you can be productive, even if you don't know any code despite designing the workflows. I see this being done in two ways: 

1. You can have someone who is non-technical design the workflow, essentially designing the control flow, specifying the input and output data, and treating this as a black box. 
2. You can have a junior developer who is familiar with the notation that Jiffy.ai is using for implementing individual execution nodes fill in the gaps. Of course, it needs some testing.

This is the development model that I see which is suitable for workflow entry. This means essentially that we don't need to engage expensive senior developers into managing the system. Also, it means that we can get from design to production faster. Essentially, this now provides an advantage, which means we use Jiffy.ai for more automation tasks as we become familiar with the UI and scripting notation.

The most valuable feature is the computer vision or OCR. That has a lot of use cases in real life. A lot of man hours can be saved, as we've seen in the finance processes and also in my future use cases. The feedback I have gotten from the team is that the OCR is quite powerful. I'm really looking forward to that.

For the finance processes, from what I know, Jiffy integrated quite well with the Oracle system. Most of the finance requests have been taken care of. It handled the integration pretty well.

I found the invoice data extraction very exciting. It is really good.

Jiffy.ai is in our cloud and very efficient. It speeds up the very time consuming, repetitive work, which would have taken up a lot of our time, if it was not automated.

It is a one stop solution for automating processes. The modular way that is assigned and works together follows a certain logic, and it encompasses a wide range of processes in a very structured and logical manner.

We did undergo some basic training on how to operate Jiffy.ai Automate. It seems to be relatively easy to implement, if you know what to do.

Their Jiffy.ai Data Interface (JDI) contains something like an audit trail of all the transactions performed. It is easily accessible, then we have an oversight of all the transactions done and the time performed as well as details of the transaction. It is pretty transparent. it is secure and hosted on our system.

With the customization option, we can write custom expressions using its compatibility with Python or other programming languages. 

Their web automation is good. It makes the developer's work easy.

Jiffy.ai integrates into existing infrastructure with a very straightforward, simple API. This was not a concern for us at all.

In the latest version, they have a solution called Docube that comes with machine learning. We have used this for the WPS processing, manually adding the keywords over the matching algorithm or things. The system automatically learns new things, and we even have an option to train the bot. This streamlines our automation process, making it easier. Otherwise, we would need one person to identify the new keywords, adding them manually.