top of page
Sales Training and Assessment

 Retail 

 Sales 

 Gen AI 

Continuous upskilling of Frontline Retail Staff via AI Generated Training Material (in multiple languages) and Assessments.

Screenshot 2024-03-27 at 12.11.09 PM.png

Problem: The client, a consumer brand, has an offline retail presence across 50+ stores in India. They need to continuously train and assess their frontline retail staff on the brand, catalog, retail sales, grooming etc. They were doing this in an ad-hoc manner i.e. sharing PDFs with low open-rates and no system to assess their staff’s learning journey.

Solution: We customised our AI Platform that empowers the Sales Head to upload an unstructured PDF/PPT/DOC (or any target knowledge base) which was fed to a LLM using RAG framework to create micro-learning training materials in multiple languages. The learning is followed by an assessment via MCQs and audio-based roleplay scenarios. The team publishes and shares the AI generated course link with their retail staff. After completing the training and quiz, a detailed assessment report is generated for the employees and the Sales Head to track. 

Demand Intelligence  Real World Events Data

 Data-as-a-Service 

 Consumer Mobility

 Machine Learning 

Data Intelligence to optimize & ramp-up driver supply based on real world events happening in a city

Screenshot 2024-03-27 at 12.21.01 PM.png

Problem: Traditionally ML teams are the right audience for a forecast-grade dataset. However, the requirement for a custom dataset came from the Growth team of an on-demand consumer mobility company. The team was launching a new city and wanted to optimize their driver supply in advance i.e. point them to locations where & when pockets of customer demand surges. This ensures high availability during the launch.

Solution: We built a custom data processing pipeline that fetches real world attendance based events (across categories: expos, concerts, conferences, sports, arts data) from multiple sources. Then, we standardised, de-duped, and filtered out spam events. Lastly and importantly, using ML, we enriched the data with 2 features (a) predicted attendance (b) end times for each event. The growth team used this continuously updated dataset to optimize driver supply during their launch in the new geo.

Real-time Last Mile Asset Tracking

 Logistics

 Streaming Data Infra 

 IoT 

Screenshot 2024-03-27 at 11.10.50 AM.png

Problem: Customer wanted to provide last mile delivery visibility to their B2B customers. There were a few off-the-shelf solutions for this. However, the team required a bespoke optimised solution.

Solution: We developed a solution based on battle tested open-source frameworks. We embedded an SDK in the mobile app of the delivery executive generating continuous real-time data. The data was cleansed, smoothened and optimised for low connectivity edge cases. Further, it was enriched with Maps APIs for calculating real time ETAs. This information was then downstreamed to a front-end JS app visible to the customer’s client. This project was optimised for reliability, cost & latency.

bottom of page