AI powered Product & Consumer Insights Solution helps an electronics manufacturer reduce 80% of manual activity


The client is a leading and fastest-growing manufacturers of headphones, earbuds and portable speakers in North America.


  • Personal audio being a competitive area, differentiated products and value propositions are imperative.
  • The existing solution built on a generic language model did not analyze the performance or review at the model level.
  • The existing tool lacked capabilities to differentiate between customer reviews, influencer reviews and posts. There was no scientific correlation between customer sentiments, influencer reviews and product sales and revenues.
  • The product and marketing team did Ad hoc analysis and created reports to make up for the gap in the existing product.


Tenzai team conducted a Purpose Driven AI workshop to brainstorm ideas with the clients’ product and marketing teams to conceptualize and design the solution.

Based on the insights, developed an NLP based product intelligence and sentiment analysis solution.

The solution had capabilities to fetch data from a wide range of sources like CeCommerce websites, social media, technology news, blogs, syndicated research through APIs and custom web scrapers.

Tenzai team trained over 10,000 documents/articles to build a custom language model to comprehend text for electronic gadgets especially headphones.

The language model differentiated between customer and influencer reviews and gave separate sentiment scores for them. It was possible to assign an overall weighted sentiment score based on features, sources and influencers.

The solution was able to attribute scores to different product features based on reviews and other content.


The solution helped business users to effectively track and/or analyze.

  • Emerging Trends and opportunities in personal audio, based on influencer posts
  • Product Intelligence based on client’s product performance and competitions based on customer reviews, influencer reviews on eCommerce, social media, news articles and blog posts.
  • Customer Sentiment based on customer emotions and sentiments from social media posts and eCommerce product reviews.
  • Customer Sentiment based on customer emotions and sentiments from social media posts and eCommerce product reviews.
  • Impact Analysis using machine learning to assess the revenue impact of promotions, social media activity, events, influencer posts of the client and competition.

Other benefits include:

  • Track 10X more data feeds and sources for research.
  • Influencer and customer sentiment increased by 12% and 18% respectively (seems incomplete).
  • Reduced manual activity by more than 80%.