Smarter.Codes Drives 23% Surge in Client Response Rate for an Ambitious US-based Tax Consultancy Firm

5 – Minute read


It consumes a lot of time for the tax consultants to review the forms and deliver tax saving strategies for clients accordingly. Going through multiple sets of documents filled with numbers and complex logical relations can be time consuming for professional service firms such as tax consultancy firms. IRS 1040 form is one such financial tax-related form. It has multiple fields within the form requiring a high volume of numerical calculations, filled with volumes of information. In addition, when the junior staff need to reach out to the senior colleagues, the waiting time for clients increases further. It impacts customer satisfaction, thus increasing the churn rate.


The AI-driven chatbot is capable of answering queries related to US tax laws in real-time. The chat-bot is trained with the latest available data, so even the queries related to the most recent news can be answered by the chatbot. 

AI would read through the complex IRS 1040 forms of the customer and offer customized tax saving strategies which are prepared in-house by the client. The tax-saving strategies are offered in real-time with a turnaround time of a few seconds versus two to three hours using manual effort, thus reducing the customer waiting time by more than 90%.

Our Approach

Training on Proprietary Knowledge

Company’s proprietary knowledge base about tax and publicly available knowledge resources provided by the client were leveraged to train the AI. In addition to it, the language learning model was also trained on the in-house tax advisory strategies that are utilized in responding to personalized tax advisory queries.

Easy Training Interface

A wikipedia like interface was developed for organizing the training data. Arranged in a structured, hierarchical database that clearly defines the concepts and relationships about tax related information. Updating the training data was as easy as updating a wikipedia page.

Document Understanding

IRS 1040 forms were used to train Computer Vision so that AI learns to read those forms just like a human does with human vision . Document Understanding AI, that incorporates elements of computer vision, was leveraged to understand IRS 1040 form, their taxonomies, their classification, and train the AI for data extraction.

Measuring success with regular upgrades

For keeping the chatbot updated, a scheduler was set up. It kept a track of the data and trained the chatbot on it to return updated answers to the user. The chatbot has recorded reduced escalation rates with increased response rate making the lives easier for the customers.


With the help of the AI-chatbot developed by Smarter.Codes, our client was able to improve their customer experience by reducing waiting times by 90% thus improving the manpower utilization rate.