quick inquiry
Case studies

Smart Search For E-Commerce

AI integrated smart search solution for ecommerce

  • Development time

    16 Months

  • Resources involved

    4 Professionals

Overview

The client wanted to have a smart search for the ecommerce through which the user should be able to perform the sentence based search like, “I want to build my home” would return all the products required for house building along with all the services available in the website.

The client also wanted to build complete solution from scratch without using any specific library or tool.

Business Goal

Develop a library which can perform the indexing and searching task from scratch.

Develop additional analyzers for sentence based searches.

Challenges

There were multiple challenges while developing the solution like,

Learning, implementing modern algorithms for parsing, processing and storing data efficiently in filesystem.

Developing independent components which fits together for whole delivering whole solution.

Achieving high cohesion and loose coupling among components.

Providing accurate results after NLP & Deep learning process.

Solution

The high level flow diagram of our proposed system is shown below.

System architecture

  • Our team started implementation parallely for building various independent components of the solution.
  • As independent components can be tested easily, it was easier for us to implement them independently, test individual component, integrate the component with the other missing pieces and then performed rigorous testing.
  • During whole solution implementation, a lot of learning was required, we did a lot of trials for various algorithms as well for the use case.
  • We developed around 12-14 different components, integrated them and tested for its full working before delivering it to the client.

Technologies

Implementation

We followed an iterative model approach to implement the solution that included the following phases:

We iterative phased approach to deliver the whole solution which required us to build 12-14 different components based on client’s requirement and tested them thoroughly for efficient usage.

NLP and deep learning methodologies were also applied on top of this solution for having best data analytics and learning from the indexed data.

We also upgraded various sub components based on modern algorithms whenever it was required or as per client’s requirement.

Road map

Requirement Analysis
  • Brainstorm
  • Innovation
  • Research
  • Resource planning
Architecture Planning
  • Architecture Design
  • System Prototyping
  • Documentation
Development
  • Front end development.
  • Front end development.
Quality Testing
  • Quality Assurance
  • User Acceptance Testing
Deployment
  • Production
  • Evaluation

Business Benefits

We delivered the unique solution to the client and as a result client organization was able to provide the sentence based search into the ecommerce platform which improved user experience and overall search performance was improved by 12%.

Got idea in mind?

Let's Talk

Contact us for your business needs anytime and our support team will be available to you 24 * 7 to answer your queries.

Related case study

Our Global Presence

Contact Us

Our Services

Hire Resources

Our Blogs

© 2009 - 2024 ProminentPixel - ALL RIGHTS RESERVED.