Drishti-Intelligent Document Processing

Lumiq's Drishti Document AI, digitize's document processing using Amazon Textract

Let's all face it, document processing manually is extremely slow and time consuming. In today's data driven world, documents represent the highest chunk of unstructured data across many industries with multiple processes and a heavy amount of document processing. Many enterprises still have trouble processing and extracting relevant information using manual or OCR bases methods which is quite expensive and unscalable.

Enter Drishti Document AI, Lumiq's complete Document Intelligence & Document Understanding engine built atop Amazon Textract. With a template agnostic, API friendly ML model, it takes negligible effort to train new documents with superior ability to extract & organize data automatically. Amazon Textract uses machine learning to instantly read and process any type of document, accurately extracting printed text, handwriting, forms, tables, and other data without the need for any manual effort or custom code.

Dristhi Document AI has an uncanny ability to extract from all types of documents and improves upon the benefits of Amazon Textract by making it easily deployable and infinitely scalable for customers. With a continuously learning AI and a rich background of proven customer use cases, it not only frees up resources internally but provides a rich end user experience as well.

DevOpsGroup logo


India, United States, Singapore, Malaysia, Indonesia


Faster Decision Making

Extract unstructured data from all types of documents swiftly & cogently to enable faster decisions

Easy Deployment & Scalability

Available as restful API deployed on kubernetes allowing infinite scalability

Reduction in Effort and Cost

97% reduction in PII redaction & 70% in data entry for human labour with pay as you go pricing model

Minimal Training required

No training required for supported classes, for others quantum of training data is relatively lesser

  • How it works
  • Pre-processing of documents can have a marked improvement on the quality of Textract output. Based on the kind of document scans and images, Drishti performs document boundary detection, orientation correction, quality improvements.

    The Drishti Document AI Engine uses a proprietary deep learning architecture that takes input not only the text extracted (from Amazon Textract), but also the interactions between text and the layout information. This is beneficial for a number of real-world document image understanding tasks such as information extraction from scanned documents.

    The Engine also uses image features such as font size, font-weight, font-family, and underlining to incorporate as much visual context as possible. This allows information to be extracted as easy-to-use key-value pairs directly from the document. Drishti document AI model keeps learning as more data is processed and progressively keeps improving the accuracy.

    Drishti also performs post-processing on the results provided by Amazon Textract, such as post-OCR correction using an advanced NLP engine and domain knowledge.

  • Key activities
  • 1) Understand Customer Requirements
    Crisp Analytics works with customers to identify & understand their pain points.
    2) Understand feature considerations
    This involves a well-architected review as well as additional workshops and desk-based assessments.
    3) Proof of Concept
    Crisp Analytics can also execute a PoC for customers to verify the solution when deployed in their environment.
    4) Deployment in client environemnt
    Crisp Analytics will deploy the solution in the client account and roll it out across all specific functions.
    5) Review with support & maintenance
    Crisp Analytics will periodically support customers with feature upgrades & support for troubleshooting.
    6) Create a future roadmap
    Crisp Analytics will work with clients to understand future requirements for subsequent enhancements to keep it scalable.
  • Customer contribution
  • Business Process Understanding

    Client needs to align right stakeholders to explain documents importance and priorities.

    AWS Account Access

    Customer must provide Lumiq personnel access to their AWS environment for the duration of the engagement.

    Security Policies and Controls

    Customer needs to provide access to internal security policies and controls for AMIs.

  • About this consultant
  • Crisp Analytics Pvt. Ltd (Lumiq) is an AWS Advanced Tier Services Partner with laser sharp focus on AI/ML solutions. Crisp Analytics is a data analytics company that powers end-to-end data-driven decision making and intelligent automation on AWS. Crisp Analytics helps businesses leverage data and drive insights into action by adopting a step-by-step iterative approach of AI/ML integration with data modeling. Crisp Analytics' 6-stage self-learning solutions cycle and holistic problem-solving approach answer business needs and enable enterprises to drive desirable outcomes.

    Crisp Analytics' deep technical expertise, pre-built functional components, and 200+ team of AWS certified data engineers and data scientists help them comprehend complex business problems and ensure desired outcomes making Crisp Analytics a reliable partner for any data transformation journey.

  • Architecture diagram

Ready to get started?

AWS Partner Highlights

Lumiq’s AWS validated qualifications, customer references, and office locations.

Explore icon
Explore all Consulting Offers

Browse our portfolio of Consulting Offers to get AWS verified help with solution deployment.

Learn more 
Build icon
Deploy a solution yourself

Browse our library of AWS self-deploy solutions to common architectural problems.

Learn more 
Find an APN Partner icon
Find an AWS Partner

Engage with AWS Partners for secure, innovative, and cost-effective custom solutions that leverage the power and scalability of AWS services to meet your needs.

Learn more