INTERNATIONAL RETAILER INDUSTRY: RETAIL
Ideating new product ideas through AI-driven trend analysis for a renowned international retailer

Ideating new product ideas through AI-driven trend analysis for a renowned international retailer

Partner Overview

Neurons Lab partnered with a major multinational, omnichannel retailer with more than 2,000 stores across 80 countries. To drive innovation in its product design, our client needed an AI solution capable of synthesizing various forms of unstructured research and trend data.

 

Challenges

This renowned international retailer is committed to innovation, evident in their teams’ design-thinking approach and focus on understanding customer needs.

To maintain their leading position, the teams want cutting-edge tools and easier knowledge-sharing, to increase collaboration and accelerate the journey from customer insight to product realization.

Broader goals include discovering more unmet customer needs, as well as enhancing the idea generation and prototyping process.

For this specific project, there were three main challenges:

  • Creative ideation: The solution had to find insights in the data to generate creative, innovative new product ideas
  • Scalability: To integrate with our client’s existing systems and allow for flexibility in the future, the solution needed to be developed on scalable cloud architecture
  • Multiple data formats: The data ingestion pipeline needed to be robust to handle multiple formats, synthesizing information into a coherent knowledge graph

 

Project Overview

diagram, ideating

We created an innovative AI-driven application designed to synthesize diverse research data and generate actionable insights. This helps drive product innovation and generate ideas for new products.

The solution identifies emerging trends and generates new product ideas based on consumer needs and market data. The application integrates and analyzes various unstructured data forms, such as video interviews, text reports, demographic and sales figures. 

To allow easy communication with the AI solution, it uses a large language model (LLM) and the MVP features interactive elements. These allow users to explore research themes and generate ideas through a user-friendly interface.

This system also supports scenario gaming exercises and offers a digital platform for ongoing access to the synthesized research, ensuring the application’s value long-term. We developed the solution on a scalable cloud architecture, to ensure flexibility for future enhancements and integration with our client’s existing systems.

Above, we have illustrated the main flows:

  1. Between the user and insight generation system. These interactions encompass submitting user requests to the customer UI and API, as well as the AI ideation process.
  2. Between the manager and data ingestion pipeline: These interactions encompass submitting research data to an internal CMS, extracting data through an API to an AI solution for processing, constructing a knowledge graph for idea generation during events, and visualizing those ideas.

 

Solution Overview 1 – Data Ingestion Pipeline

For the data ingestion pipeline, research data – including text reports, video interviews, sales figures and demographic data – is submitted to an internal CMS for AI-led processing, creating a knowledge graph. 

This is the data collection or ingestion process to construct the knowledge graph. The solution ingests the data, reprocesses it, extracts triplets, stores them, and indexes them in a knowledge graph.

We store the data in Amazon S3 and use Amazon Transcribe as part of the step function:

diagram, ideating

We designed the data ingestion pipeline solution to effectively process and manage various types of input data sources. 

This converts them into a structured format suitable for analysis and storage in a graph database. Therefore it generates relevant ideas without AI hallucinations.

aws architecture, ideating

Solution Overview 2 – Ideation

The ideation solution is structured to enhance the generation of innovative product ideas through a highly interactive and responsive system. 

The ideation solution leveraged the AWS Fargate API layer, LangChain LLM framework and a Neo4j graph database to generate innovative product ideas, communicated to users via Anthropic Claude:

diagram, ideating

Additionally, we stored search logs into the private AWS Redis database, for further analysis of user requests and to make further optimisations. 

Let’s say the user-specified theme is around sport equipment – asking what is trending, what the latest versions are, and so on:

  • The solution processes the query, then the Knowledge Graph search retrieves data and generates new ideas
  • If it’s not enough information yet, then the solution searches for additional nodes in the graph to find some additional data, generating better answers, before providing several responses to the user

AWS Fargate API Layer, ideating

Results

The solution:

  • Generates new product ideas based on consumer needs
  • Supports scenario gaming exercises
  • Identifies emerging trends based on market data
  • Creates actionable insights from unstructured data
  • Runs on scalable cloud architecture to support future enhancements and integration with other systems.

The solution helps this renowned international retailer to identify emerging trends and explore new product ideas based on consumer needs.

For the data ingestion pipeline, we delivered a robust data ingestion pipeline that can handle multiple data formats and synthesize information into a coherent Knowledge Graph, including:

  • Data ingestion pipeline and infrastructure deployment code
  • Knowledge Graph based on client data
  • Testing reports

We also developed an ideation component utilizing insights from the data to generate creative product ideas, including:

  • Ideation and infrastructure deployment code
  • API documentation
  • Testing reports

Potential future use cases for the solution include support for customer journey mapping, behavior insights, trend influence mapping, sentiment analysis, predictive analytics, and product recommendations for customers based on their preferences.

 

About Neurons Lab

Neurons Lab is an AI consultancy that provides end-to-end services – from identifying high-impact AI applications to integrating and scaling the technology. We empower companies to capitalize on AI’s capabilities.

As an AWS Advanced Tier Services Partner and Generative AI Competency holder, we have successfully delivered tailored solutions to over 100 clients, including Fortune 500 companies and governmental organizations.

Our global team comprises data scientists, subject matter experts, and cloud specialists supported by an extensive talent pool of 500 experts. Get in touch with the team here.

Get in touch Leverage the power of AI for your business.
Personalizing promotions and product recommendations in retail with AI-powered models
Multi-channel retailer: Personalizing promotions and product recommendations in retail with AI-powered models
Creating an AI-driven video analysis platform with audience personalization data for Zentiment
Zentiment: Creating an AI-driven video analysis platform with audience personalization data for Zentiment
Detecting road damage to prevent driving hazards using predictive AI for ASH Sensors
ASH Sensors: Detecting road damage to prevent driving hazards using predictive AI for ASH Sensors