Challenge

The client is a leading global Tobacco Conglomerate looking to test their new product, an e-cigarette prototype prior to launch in the market.

Solution

The approach involves Face and Sentiment analytics using NLP methods to quantify and rate various prototypes on Spontaneous Reactions, Sensorial Reactions, Lifestyle Implications and Areas of Improvements.

The CAInatics tool calculates LSTM generated sentiment scores – with predictive model accuracy average on random dataset (compared to manually audited data) which is more than ~ 88% (the current state-of-art for unsupervised sentiment assignment ACCURACY is more than ~ 90%) Custom neural net generates word embeddings (Assignment of 100 dimensional vectors to each of the words) Face Analytics on set of Users (Sample) basis below concepts Valence on positive or negative affectivity Arousal measures on how calming or exciting the product is.

Impact

Score-Card on Prototype Generate with scores basis Sentiment analytics and Face analytics data-set considering.

Challenge

The client a leading CPG brand was looking to assess the impact of Consumer Brand Promotions in a Retail Store.

Solution

We deployed the CAInatics Computer Vision based analytics solution to monitor and measure:

• Footfall | Demographics
• Movement in-store during store operation timings
• Engagement with in-store Consumer Promotions
• Brand/SKU/Meta-data based in-store consumer movement tracking

Impact

Score-Card on Prototype Generate with scores basis Sentiment analytics and Face analytics data-set considering Spontaneous Reactions, Sensorial Reactions, Lifestyle Implications.

Video

• https://youtu.be/vmTxVDkKD9g

Challenge

The client is a Fortune 500 company, a leading Consumer Goods player with a large presence in brick-and-mortar retail.

The company was using human-based processes to track, trace and fulfill their CPG/FMCG products/brands/SKUs at several brick-and-mortar retail distribution point-of-sales. Given the scale of operations on various SKUs/brands/variants handled at each point-of-sales touch points, the process of acquiring information was slow and unreliable, leading to missed sales opportunities and absence/delay in customer satisfaction.

Solution

Our Point-and-shoot camera-based AI mobile application makes this process simpler, reliable and easy to integrate with existing retail sales operations. The entire process can be executed within minutes to capture images and process image analytics to get real-time and accurate data reporting to ensure best retail execution with relevant information to make faster decisions.

Impact

Automated measurement and monitoring of KPIs like Distribution, Shelf-Share%, Price-Points, Competitive Intelligence and Planogram Execution/Compliance in real-time led to 20% improvement at points-of-sale.

Video

• https://youtu.be/6vVe3GNVTE8

Challenge

The client CavinKare is a pioneer of sachets in India. 80% of their hair care business comes from sachets. They wanted to know their share of visibility in the kirana outlets.

Kirana outlets are very small spaces, cluttered with several products hanging on a wire across the shop.

The sachet sizes are about 2”x1.5”. The challenge was to count competitive sachet strips and provide as much granular details as possible.

Solution

Using the image recognition expertise built into CAInatics, Constems-AI engineers trained the data on this smallest SKU in the shampoo category not only to read brand names but also sub-categories, ingredients and benefits.

Outcome

This is a technological breakthrough, a never been done before feat.

Client Industry

Supply-Chain | e-commerce | Consumer Products | Consumer Electronics

Challenge

Damaged product packages reaching the end-consumer from the Warehouses. Additionally, wrong products dispatched to end-consumers from Warehouses.

Solution

Our integrated inspection system, based on AI Computer Vision can create a system for 100% digital inspection and visibility of shipments at the time of receipts/ dispatch w.r.t. Count, Barcode/ QR code detection, SKU type detection and Box damage condition/ detection The system is enabled with HMI and Tower Lamp for alerts & escalations and integrated with Warehouse Management System (WMS)

Impact

Zero Tolerance on Defective packages leaving Warehouses and 100% Verified dispatches was possible with our solution.

Video

• https://youtu.be/An9yObGNUpo

The client is a Fortune 500 company and a large contributor to the F&B category in India

Challenge

Generating New Product Development Idea using Consumer interaction (Images & Video) data.

Solution

Videos and Images collected from consumers on their daily life experiences. Deep Learning model applied on Image Processing → Feature Extraction → Feature Selection → Classification → Output (Findings) on preferred Consumer Patterns.

Impact

The client used this information to arrive at Top (#3) New Product Development Ideas for Testing with End-Consumers.

The client is a leading FMCG company and major player in the Food and Beverage space

Challenge

The company was looking for insights on Consumer Trends / Patterns in New Markets for Mothers and Teen Children.

Solution

Data mining from social media was the adopted approach. Using Open APIs of Twitter and Instagram a total of 9,500 Teen profiles and 8,700 Mother profiles were shortlisted to analyze their outdoor eating patterns. Further analyses was done using Image Processing, Sentiment Analytics and Classification using Unsupervised Machine Learning. Videos and Images collected from Open APIs from social media were mined for classes → #1. Vegan, #2. Drinks #3. Regular Meal with Family.

Impact

The client was able to use the findings/insights generated by CAInatics to devise their Market Entry Strategy.

Client Industry

Cool.AI Smart Visi-Coolers primarily cater to refrigerated perishable goods produced by Food & Beverage companies, wherein they can monitor the inventory and status of products inside the fridge.

Challenge

Provide insights into perishable stock levels, identify potential usage of cross brand products and optimize inventory management.

Solution

Our Cool.AI Smart Visi-Cooler features portable cameras equipped with AI technology, capable of capturing and GPS-tagging images. These images are sent to the cloud at scheduled intervals. Our tailored AI-driven Machine Learning algorithm processes images and metadata, extracting pertinent details of product SKUs within the Visi-coolers to ascertain stock availability.

Impact

• Tracks the fridge(Visi-Coolers) locations to alert clients to intervene where consumption levels are insufficient.
• Provides consumption visibility from the fridge (Visi-Coolers) kept at the retailer location to predict stockout.
• Provides valuable insights into brand level stocks by identifying potential usage of cross brand products to ensure Visi-cooler purity.
• Tracks movement and handling of fridge (Visi-Coolers) to prevent loss due to theft or carelessness.