What is our primary use case?
We are primarily using it for processing documents, particularly for procurement and payroll. We have a case for HR to deliver a document for attestations for people who are leaving the organization. They are provided with an attestation letter. We are primarily using it for payroll and procurement, so far.
We have our own virtual servers where we host on our own servers, but they are virtual and not physical. The machines running the bots are virtual machines as well.
How has it helped my organization?
We have not been using it for anything other than the standard automation process, which performs well.
In terms of the accuracy, it has done particularly well. For staff, it is common when doing repetitive work that there tends to be mistakes, even if you are familiar with it. There will be a missing digit or letter here and there, but Jiffy.ai Automate never does that. It is accurate to a "t". It encompasses errors only if the input is poorly done, so it does not commit errors when it does this process.
Jiffy.ai Automate has reduced manual processing for the 25 percent of the processes that it has automated. We have many processes, not just one, that Jiffy.ai Automate processes. Each process takes between five to 10 minutes, depending on the network. It could be as fast as two minutes if you did it manually, but then it also depends on how the network performs, as it may be inaccurate as well. Jiffy.ai Automate may take slightly longer, because of the way it works (usually five minutes). If the network is a bit slow, then maybe it takes at most 10 minutes and you need to times that by 2000 transactions minimum. In addition, there were certain months when the numbers of transactions that Jiffy.ai Automate processed spiked, so we hit as high as 5,000 transactions.
The solution was proven to be especially essential during the pandemic. For example, the work that Jiffy.ai Automate does is around 25 percent of the data process. If Jiffy.ai Automate had not been operating during the pandemic, that 25 percent would probably have taken a much longer time than it normally does. Though, it depends on working from home conditions and some other factors. If it hadn't been for Jiffy.ai Automate, then the staff would probably have only got 30 percent done. The turnaround time would have been much longer; it could have been at least two to three times longer.
It is important that Jiffy.ai Automate does its work, so payments will be initiated and processed on time to our vendors, partners, and suppliers.
We engage Jiffy.ai to do reusing, modifying, and building process automation work for us. We only inform them what we want them to do, then their staff will follow up closely and do the coding, then the actual automating process. We explain the process to them, they understand it, and then they capture it in their system. They run it, test it, and then confirm that it works with us. We are practically hands-off with it.
What is most valuable?
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.
What needs improvement?
They are doing part of the automation process, but it is not entirely end-to-end. They are filling in the gaps for us, and from an end-to-end perspective, they fulfill part of it. They breakup about 25 percent, then 75 percent is actually work done by the staff. However, for the 25 percent that is required, this is actually very time-consuming, especially now during the pandemic. Forwarding documents from home is a pain, as it depends on individual staff and their Internet connection as well as the performance of their ISP. So, we have to take that into account.
I have a background in IT and I have an understanding of how the RPA system works. If I am programming this, I know it will be a hassle. It is mostly graphical, then something that you just need to type. As long as you know where to press and what to do as well as have some basic programming skills, it is quite easy. The solution has just not closed the gap of being accessible to non-IT users. If you are a non-IT person, then this all looks like gobbledygook. Maybe that is something that can be improved upon.
For how long have I used the solution?
We have been using it since the end of 2018, between November and December.
What do I think about the stability of the solution?
Thus far, it has been pretty stable. Its only bottlenecks are due to network performance, or even the issues that we face with IT sometimes. For example, the solution runs on virtual machines, and sometimes those machines don't start properly because there are some issues with logins, etc. Mainly, that is beyond their control. The base system itself is quite stable.
It fits our requirements well. There are no issues whatsoever that I can think of. Jiffy.ai Automate understands the way that our systems work. The process that they are doing involves several systems: the repository system, the ERP system, our SharePoint, and our single sign-on system. It works across all these various systems and their operations well, adapting well to the environment.
So far, it has been reliable, except for once or twice where there could have been some technical issues involved. However, the logic and processing of the system are quite solid.
Unfortunately, because of the pandemic, we have been stuck for a while. We had to adjust our upgrade to version 3.2. However, due to the pandemic, some infrastructural changes will not be able to be done. Therefore, we are still stuck with version 2.7. We intend to upgrade to version 4.0 this year.
What do I think about the scalability of the solution?
In the beginning, we had three processes. Then, we gradually increased it to seven. There have been no issues whatsoever with scalability. So far, it has been working fine.
We have between 2000 to 5000 transactions processed by Jiffy.ai Automate every month.
Jiffy.ai assists 50 of our staff, who are part of the procurement and payroll teams, to do their work.
How are customer service and technical support?
Thus far, their support has been great. They are quite responsive. If there is something that requires an action from our end, they understand our process flows very well.
How was the initial setup?
It took us around two weeks to implement the first process. So, it was quite fast. Then, we did some verification and validation, then it was up and running quickly. They understood our needs very well and had us up and running within two weeks for the first process, which was quick. Subsequent processes took another additional two weeks.
What about the implementation team?
There were only two staff from our company in the beginning, my colleague and myself. I did the procurement part and my colleagues did the payroll part.
Jiffy.ai did everything from the coding to the support of the system. We were the ones telling them the process. We were also the liaison between them and the business unit. Jiffy.ai codes the process, then they execute, run, maintain, and support it.
Our experience with their setup support was satisfactory. Out of 10 times, there were maybe one or two lapses. This was because of possible miscommunication or they could not answer in time due to some factors. However, most of the time, they were quite responsive.
What was our ROI?
In terms of the processing of the transaction, Jiffy.ai Automate scores monthly between a 95 to 98 percent success rate for all transactions. The other two to five percent are failures, mostly due to technical issues or an error caused by the staff, e.g., if they entered a wrong number. Then, it becomes an error because the number doesn't correspond to that transaction. It could also be a network outage or system issue.
Jiffy.ai Automate does the work of around three of our normal assistant level staff doing data processing. Each of those staff probably makes $20,000 to $30,000 per year.
What's my experience with pricing, setup cost, and licensing?
The licensing is not expensive; it is quite cheap. The expensive part is to support and maintain it. Understandably, we engage them to do a lot of work that could have been done by our own staff, if we had an RPA team. I wouldn't hold that against them. It is just that the cost of support is on the higher side for us.
If there is an upgrade, they will also charge us for the man-hours required to perform the upgrade.
Which other solutions did I evaluate?
We interviewed a few suppliers and tested some of them. Then, we ended up with Jiffy.ai. In 2018, we did talk to the big companies, like UiPath, Blue Prism, and Automation Anywhere. The problem was our center is located in Malaysia, South East Asia, so there weren't many partners available locally at that time. The way that they were going to implement, they would have to look for a partner somewhere in Singapore to come over to our center and perform the implementation. That was a bit of a hassle because there weren't many RPA partners then. Now, before the pandemic, there were all these new partners popping up in the market, growing in 2019. However, back in 2018, Jiffy.ai was the only one whom we could use without going through a middleman. They were the only ones willing to come to our office, sit with us for a few weeks, and analyze the system as well as the process. Then, they got everything up and running. They were the only ones willing to do that, which is the primary reason why we selected them.
Jiffy.ai is not as mainstream as the major players: Automation Anywhere, Blue Prism, and UiPath. Those three are constantly on people's minds when they think of automation. Other companies, like Kryon RPA, have hit the nail on the head in terms of marketing and advertising. Jiffy.ai doesn't have that type of exposure and could stand to have more visibility.
Other companies market, "Even a dummy can operate the bot by themselves," and Jiffy.ai doesn't. Maybe, they are also being honest. However, they are not marketing themselves like other companies, where even a guy without IT knowledge can operate that bot. However, all these companies which say, "Even dummies use it," neglect the fact that you will not be afforded some granular controls.
What other advice do I have?
I learned that processes can be made more efficient with automation. That is a no-brainer. However, I feel that there are still a lot of misconceptions about the role of bots and people are afraid of them. They are afraid of losing their job due to a robot. RPA is something like a sneaker that allows you to move faster and do more. It is not a replacement for a person. A person is still needed to do the work for the direction you want to go and the target that you need to reach. A bot allows you to do things faster.
Give them a try, then you can compare them to the other solutions. I think you will be surprised to find that they are quite responsive and understanding of your automation needs. They grasp things quite quickly. All the software automation in the market basically does the same function. What makes the difference from them is the human factor.
We have not used the advanced features. They have things like machine learning and some AI thrown in, but we have not used them. I find that Automate is basically just a descendant use of Jiffy.ai itself. Also, we are not directly programming on it.
Most of the time the process is hands-off and we don't interfere with what Jiffy.ai Automate is doing. When it encounters an error, due to some human error or a system issue, then we will have to go in and perform manual intervention.
We are upgrading our ERP and are unsure if Jiffy.Ai Automate is compatible. If it is compatible, then we will use their solution. We are open to new systems. So, if a better system comes along, then we might consider it.
I would rate the solution as an eight (out of 10).
Which deployment model are you using for this solution?
Which version of this solution are you currently using?