HAVE A QUESTION?
Call us : +1-945-209-7691
- Software DevelopmentIT Consulting & DesignAI & Data SolutionsQuality AssuranceTeam & Resource SupportBusiness Support Services
EPS (Email Processing System) is an AI-driven automation platform developed for Echem to intelligently process and manage high volumes of quotation emails in the chemical trading domain.
Echem partnered with us to build the Email Processing System (EPS), an intelligent platform to automate the intake, classification, and data extraction from high volumes of quotation-related emails in the chemical trading space. The system fetches emails from multiple mailboxes, classifies them using machine learning models, and extracts key data like CAS numbers, chemical names, and quantities, even from complex tables within emails. EPS integrates fine-tuned large language models (LLMs) for advanced unstructured data extraction while providing a user verification interface to review and correct data before pushing it to Echem’s internal request modules. This reduces manual workload, speeds up quotation processing, and ensures accurate, consistent data for procurement and follow-ups. The scalable architecture allows Echem to extend automation to invoices, order confirmations, and payment tracking, future-proofing their operations while maintaining data integrity across ERP and CRM systems.
Echem is a domain expert in chemical trading and sourcing, managing a high volume of client and vendor interactions primarily through email.
To scale operations and reduce manual processing, Echem initiated the EPS project to automate the intake, classification, extraction, and management of quotation-related emails, ensuring faster and more accurate responses.
Chemical Trading, Supply Chain
Email Automation System using ML & NLP
Comprehensive solutions designed to enhance user experience and drive business growth.
Sorts incoming emails by category, priority, and business relevance using machine learning models.
Extracts CAS numbers, chemical names, quantities, and other key fields from quotations, including from complex tables.
Uses a custom fine-tuned Mistral-7B LLM to enhance contextual understanding and extraction accuracy.
Enables users to review, confirm, and edit extracted data and email labels with an intuitive web interface.
Provides real-time insights on email volumes, processing status, and label accuracy.
Allows configuration of custom rules for email classification and data extraction logic to align with evolving business needs.
Fetches and centralizes emails from multiple customer and vendor mailboxes into a unified system.
Pushes verified, structured data into Echem’s EMS Quote modules, with future readiness for ERP and CRM syncing.
Tracks all user actions, errors, and label corrections for transparency, auditing, and continuous model improvement.
We identified key pain points and developed targeted solutions to transform the resort's digital presence.
Managing high volumes of emails with mixed structures (plain text, HTML, and embedded tables) presents challenges.
I must distinguish between similar email types such as quotes, inquiries, and follow-ups.
I am responsible for extracting highly variable chemical data, such as CAS No., chemical names, and quantities, from semi-structured emails.
I am responsible for ensuring data accuracy and validation with minimal manual intervention.
We ensure data consistency by integrating with external ERP and CRM systems.
Implemented AI-based email classification models trained on historical labeled data to categorize emails efficiently.
Developed a rule-based label correction engine to refine AI predictions and adapt to evolving business needs.
Integrated fine-tuned local LLMs to extract chemical-related data from unstructured and structured content.
Built a custom module for extracting table attachments.
Designed a web-based verification interface for users to validate and correct extracted data seamlessly.
Enabled robust logging and exception tracking for transparency and easy debugging.
Established integration pipelines to sync verified data with Echem’s EMS Quote module and prepare for ERP/CRM connections.
Visual highlights showcasing the transformation and key features of the new website.
echem-epsechem-eps
echem-eps
echem-eps
echem-eps
echem-eps
echem-epsLet's discuss your project and create a custom web application that drives your business forward. Get started with a free consultation today.
" "
70–80% reduction in manual email triaging and labeling, freeing up valuable staff time.
70–80% reduction in manual email triaging and labeling, freeing up valuable staff time.
Faster, more accurate quote processing, improving turnaround times and customer responsiveness.
Minimized human error in data entry and quotation processing, ensuring data reliability.
Continuous learning loop through user feedback, enhancing model accuracy over time.
Scalable architecture ready for future extensions like invoice and payment tracking automation.
Integration-ready structure for seamless syncing with ERP and CRM systems, supporting end-to-end digital workflows.
Improved operational efficiency and cost savings by automating repetitive, high-volume tasks.
USA
Singapore
Switzerland
India