14 Feb 2024
/
Fine-Tuning Large-Language Models
Automate Contract Classification with Large-Language Models (LLMs)
The challenge
Imagine a team, dedicated with the task of manually sorting through around 400 contracts every day, each needing to be categorized into more than a thousand specific categories. This process was not only time-consuming, but also prone to human error. That's where we came in.

Time Saved
350
Hours/ Month
Cost Saved
15.000€
/ Month
# Contracts
8.000
/ Month
The NeoTask Solution
We crafted an automated system that classifies the contracts into over a thousand activities. This approach involved fine-tuning Large Language Models on the client's contract data, ensuring a deep understanding of the nuances involved in each document. Recognizing the importance of reliability, we introduced a trust metric into the system. It allowed us to calibrate the model's predictions, giving our client a clear view of how certain the system was with each classification. This not only enhanced the accuracy but also offered our client a transparent insight into the system's performance.
Integration and Impact
Seamless AWS Integration: The solution was smoothly integrated into the client's existing AWS workflow, facilitating immediate adoption.
Annual Efficiency: Now processes approximately 90,000 contracts annually, saving over 4000 hours of manual work each year.
14 Feb 2024
/
Fine-Tuning Large-Language Models
Automate Contract Classification with Large-Language Models (LLMs)
The challenge
Imagine a team, dedicated with the task of manually sorting through around 400 contracts every day, each needing to be categorized into more than a thousand specific categories. This process was not only time-consuming, but also prone to human error. That's where we came in.

Time Saved
350
Hours/ Month
Cost Saved
15.000€
/ Month
# Contracts
8.000
/ Month
The NeoTask Solution
We crafted an automated system that classifies the contracts into over a thousand activities. This approach involved fine-tuning Large Language Models on the client's contract data, ensuring a deep understanding of the nuances involved in each document. Recognizing the importance of reliability, we introduced a trust metric into the system. It allowed us to calibrate the model's predictions, giving our client a clear view of how certain the system was with each classification. This not only enhanced the accuracy but also offered our client a transparent insight into the system's performance.
Integration and Impact
Seamless AWS Integration: The solution was smoothly integrated into the client's existing AWS workflow, facilitating immediate adoption.
Annual Efficiency: Now processes approximately 90,000 contracts annually, saving over 4000 hours of manual work each year.
14 Feb 2024
/
Fine-Tuning Large-Language Models
Automate Contract Classification with Large-Language Models (LLMs)
The challenge
Imagine a team, dedicated with the task of manually sorting through around 400 contracts every day, each needing to be categorized into more than a thousand specific categories. This process was not only time-consuming, but also prone to human error. That's where we came in.

Time Saved
350
Hours/ Month
Cost Saved
15.000€
/ Month
# Contracts
8.000
/ Month
The NeoTask Solution
We crafted an automated system that classifies the contracts into over a thousand activities. This approach involved fine-tuning Large Language Models on the client's contract data, ensuring a deep understanding of the nuances involved in each document. Recognizing the importance of reliability, we introduced a trust metric into the system. It allowed us to calibrate the model's predictions, giving our client a clear view of how certain the system was with each classification. This not only enhanced the accuracy but also offered our client a transparent insight into the system's performance.
Integration and Impact
Seamless AWS Integration: The solution was smoothly integrated into the client's existing AWS workflow, facilitating immediate adoption.
Annual Efficiency: Now processes approximately 90,000 contracts annually, saving over 4000 hours of manual work each year.
14 Feb 2024
/
Fine-Tuning Large-Language Models
Automate Contract Classification with Large-Language Models (LLMs)
The challenge
Imagine a team, dedicated with the task of manually sorting through around 400 contracts every day, each needing to be categorized into more than a thousand specific categories. This process was not only time-consuming, but also prone to human error. That's where we came in.

Time Saved
350
Hours/ Month
Cost Saved
15.000€
/ Month
# Contracts
8.000
/ Month
The NeoTask Solution
We crafted an automated system that classifies the contracts into over a thousand activities. This approach involved fine-tuning Large Language Models on the client's contract data, ensuring a deep understanding of the nuances involved in each document. Recognizing the importance of reliability, we introduced a trust metric into the system. It allowed us to calibrate the model's predictions, giving our client a clear view of how certain the system was with each classification. This not only enhanced the accuracy but also offered our client a transparent insight into the system's performance.
Integration and Impact
Seamless AWS Integration: The solution was smoothly integrated into the client's existing AWS workflow, facilitating immediate adoption.
Annual Efficiency: Now processes approximately 90,000 contracts annually, saving over 4000 hours of manual work each year.
Read More Articles
We're constantly pushing the boundaries of what's possible and seeking new ways to improve our services.