To summarize, we talked through the pain points of:

- Manually pulling down documents from Salesforce
- Identifying which contracts were required for review
- Manually reading the same contract terms over and over to “check the box”
- Documenting review and revenue accounting assessments manually
- Issues with communication among reviewers
- Lack of consistency in review results
- Lack of a single place to retention information and documentation, and
- Re-review and preparation for audit
‍

We always knew that these problems don’t just come with one solution. Having an AI contract accountant that can solve these problems through automation is not enough. We knew that the front end would have to function to the specifications of a super technical accounting team. We knew that the back end would have to be seamless; taking into consideration differing volumes on a monthly basis, the flow of data from up-stream systems to down-stream systems, and everything in-between. To get some insight into the focus it takes on the technical side, we talked to a couple members of Klarity’s high performing engineering team.

For the front end, we talked to Charlie Dieter, Klarity’s Senior Frontend Engineer to get his perspective on what it makes to make a great product:

We always knew that these problems don’t just come with one solution. Having an AI contract accountant that can solve these problems through automation is not enough.

Charlie said, “For product and user experience design, nothing beats actual user testing and taking lots of notes. When it comes to ingesting customer feedback, the focus tends to lean toward what is requested most frequently by the most customers; it’s dynamic. When we hear the same challenge multiple times, we know that’s impactful information, and we use it to make strategic product decisions. RevRec processes have similarities, but also some variances between our customers. Creating a solution that will work for many different companies, so the engineering work is not repeated, is the biggest challenge for me. Thinking through the root cause of our customer’s challenges will help us scale and help more companies streamline their process. At Klarity, we focus on quality, but also want to move super fast, so we follow the steps of a Design Sprint where we rapidly build designs and rough prototypes and get user feedback as quickly as possible, and iterate; rinse and repeat. Once most customer feedback is positive, we know we are on the right track and can invest time building the full fledged product.

Enterprise software is built to solve business problems, but it’s ultimately used by people. People ’taste with their eyes first’. Even if the technology has great value, if it looks too rough, some people will be immediately turned off. So you have to balance the time investment in creating high fidelity designs against the drawbacks of potentially looking unpolished.

In software development there is a practice called CI/CD (continuous integration, continuous deployment) that basically means you should automate building/testing/deploying. What that means is that I can write code and immediately push changes to a development environment that mirrors what it will be in production. This helps us stay on the same page with product owners, clients, and other stakeholders.

Easily creating development environments, beta environments, and onboarding users requires a lot of infrastructure. Having talented backend people makes it so you can iterate more quickly through the design and prototype process.”

To give us some insights on the backend, we spoke to Klarity’s backend Software Engineer, VenuMadhav Kattagoni. He said, “The complex backend systems which make the core user experience possible are equally important. Ensuring that backend systems are up to speed with the changing product design requirements is very important for agility. Klarity employs various backend systems like datastores and deep learning models to extract various pieces of information important to our customers. Availability, scaling and runtime of these systems are important to ensure smooth user experience. Alongside, security is extremely important when working with customer data. To ensure security, we employ all the recommended security standards throughout our systems. When ensuring the best security practices, the speed of development takes a hit. Balancing both is extremely important. This requires ensuring attention to every minute detail, while architecting and developing systems.