Written by Enoch (Founder/CEO of Soraban)

We’re in an exciting times but at the same time, I must also say that it’s an incredibly stressful time for everyone in accounting (including software provider like us!) as we expect a lot of changes in the job market for accounting. There will be winners and losers. General consensus seems to be that AI will not exactly replace jobs but professionals who use AI will replace traditional professionals. This is a great news for people starting out, and terrible news for people who pride themselves in their positions.

I’ve been fortunate to have been a software engineer for several year before running my own tech-enabled accounting firm for two years to give me just enough taste into what’s possible. As the Founder of Soraban, I lead the product development and currently working closeyl with Open AI team and been able to see what’s possible from the forefront. This article is put together today (10/21/23) so much of these may get outdated in the coming months.

Once you step backwards a few steps from all the hype, you realize while there are quite a few ways you can find value today, but it’s no mean you will be 10X more productive today from using AI.

It's no understatement to say that AI has captured the imagination of the public. Just consider how quickly it has been adopted.

We’re talking about 100 million users! That’s incredible.

Let's explore some of the other areas where AI is currently making a difference that I believe that can actually provide value.

1. Reformatting: AI can easily convert data from one format to another, such as Excel formats.
2. Categorizing Bank Transactions: AI can accurately categorize bank transactions, saving time and reducing errors.
3. Automating Email Communication: AI can draft client emails, particularly in difficult situations, ensuring professional and effective communication.
4. Chatbot-like Documentation: AI-powered tools can generate documentation that resembles a conversation with a chatbot, providing clear and concise information.
5. Putting together marketing assets

Jason Staats (from Realize group) put this incredibly well together in this playlist as well.

For example, AI created the below phrase just based on “could you put together a 1000-word blog post on…. and mention X Y Z”. Difficult to tell whether an AI put that together or human.

Wow.

How about putting together client emails?

“You are an accountant. Could you put together an email to let client know that he still owes me $1500 for the individual tax return my team finished last week. Please paymet by this week. Say it nicely?”

“You are an accountant. Could you put together an email to let client know that the work they are asking is out of scope?” and you will get something like below:

There are many behind-the-scenes applications of AI that are currently helping companies in various ways, including some of the work Soraban is doing. Tax return reading, for example, has always been a challenge. While there are tools available that claim to read tax returns well, none of them have achieved perfection yet. Some areas where AI is making an impact include:

• Understanding complex areas such as K1 forms and supplemental information.
• Extracting details that template-based optical character recognition (OCR) may miss.
• Achieving human-level text extraction and understanding.
• Mimicking human-like responses and dialogue.
• Recognizing and interpreting different situations related to payment vouchers.
• Generating summaries based on calls, with tools like Fathom.video, https://fireflies.ai/, Gong, Chorus performing well in this area.

It's important to note that there are still some areas where AI struggles, particularly in what we call "hallucination." As Zapier blog posts explains in https://zapier.com/blog/ai-hallucinations/, hallucination is when an AI says something inaccurate with confidence that it’s true. That’s obviously no good in taxation.

I would also add that most tools you try out in AI will have a lot of gimmicks too. Probably close to 90% of the things you test out, but I assure you there’s 10% of use cases that work incredibly well.

Being part of the esteemed Y Combinator network, we share ties with the pioneering OpenAI team, granting us exclusive access to breakthrough technologies. Join hands with Soraban and play an instrumental role in sculpting the future of accounting. If you are interested in testing out above, reach out to us and depending on your needs, we’ll give you early access to what we’re building!

‍

References

• Jason Staats' playlist: YouTube
• Zapier blog post on AI hallucinations: Zapier Blog
• Fathom.video: Fathom.video
• Fireflies.ai: Fireflies.ai
• Y Combinator: Y Combinator
• OpenAI: OpenAI