Device - Macbook Air

Challenge / Problem Statement

The credit management ecosystem faced multiple challenges:

  • Manual tracking of credit card bills often led to missed payments and penalties.
  • Customers lacked a unified platform to consolidate and manage their bills.
  • Extracting structured payment details from unstructured data (emails, PDFs, SMS) was complex.
  • Lack of visibility into spending patterns reduced customer engagement and financial awareness.
  • Personalized insights and notifications were missing.

Objectives

  • Automate extraction of credit bill details from multiple sources (emails, SMS, PDFs).
  • Provide a single platform to track, manage, and pay bills.
  • Notify customers with real-time alerts and reminders for due payments.
  • Enhance customer engagement with rewards and referral features.
  • Offer data-driven insights to improve customer financial management.

Process & Implementation

  • Built a data flow pipeline to extract unstructured bill data using OCR.
  • Applied ETL processes to clean and transform the extracted data into a warehouse (BigQuery).
  • Integrated visualization and BI reporting using Looker.
  • Leveraged MoEngage for customer engagement, notifications, and rewards automation.
  • Designed ML algorithms to analyze user trends and enable personalized marketing.

Our Solution

We engineered a credit management platform (Cheq One) that:

  • Extracts data directly from emails, SMS, and PDFs.
  • Calculates payable amounts and automates bill reminders.
  • Provides a unified dashboard for bill tracking and payments.
  • Includes referral and rewards systems to enhance engagement.
  • Delivers BI-driven insights for both customers and stakeholders.

Tools & Tech Used

  • Backend: Python
  • AI/LLMs: Gemini 1.5, GPT-3
  • Retrieval: FAISS, RAG pipeline
  • Embeddings: Hugging Face / Sentence Transformers
  • Automation: REST APIs for CRM integration

Results & Impact

  • $10M+ investment raised due to product’s innovative data extraction and engagement features.
  • Significantly improved customer adoption and retention.
  • Increased payment on-time rates with proactive reminders.
  • Enhanced customer segmentation and targeted marketing.
  • Improved financial visibility for customers.

Key Takeaways

  • Automating unstructured data extraction unlocks massive opportunities in FinTech.
  • Combining ML + BI improves customer engagement and decision-making.
  • Rewards and referral systems increase customer stickiness.
  • Scalable pipeline ensures adaptability across financial domains.

Project Duration & Team

  • Duration: 3-4 weeks
  • Team:1 AI Engineer, 1 Backend Developer, 1 Data Analyst
Share