Dr. Animesh Deval

• SuperApp Analytics (Ongoing)

Understanding the requirements from cross-functional teams such as product team and marketing team

Identification of relevant KPIs to derive customer insights

Development of analytics to derive insights such as Total app installs/registration/uninstalls, Total active base, Total paying base, Total daily sessions, Average session time, Social media share, Average revenue per user, Top 10 services on the App, Customer retention rate, Customer lifetime value, Conversion rate, Repeat purchase rate etc at SuperApp platform level and individual service level

Development of music recommendations system - popularity, content-based and collaborative filtering

• SLA Management Analytics

Developed and automated python scripts for the analysis of third parties responsible for SLA breach for the data stored across 54 servers.

Total 12 dynamic scripts were developed which could detect 30+ request and response patterns.

Detailed analysis showed ~ 0.1% transactions breached 3 sec response time.

The analysis helped the telco client to pin-point the problems present at the third-party to rectify the same to enhance their user's experience.

• Development of Scalable Churn Management Module (CMM)

Innovative module by integrating 4+ machine learning models - Churn Model, Survival Analysis, Usage Segmentation, RFM (Recharge) Segmentation and Drop in Usage Automation of 4+ models for easy of use to clients through UI

Conceptualized CMM UI Layout for 4 modules Designing of Visualize & Explore platform for enhanced insights and reporting by the client

Designing of Build Campaign Use-case platform to encode customers based on use-case logic. This was achieved by Assistive Builder - most common use-case template and Custom Builder - advanced interactive visualization to help build custom use-case

• Development of Recommendation Engine for telco prepaid

Developed and automated three recommendation systems to overcome cold start problem of new prepaid product and/or new customer viz. User based collaborative filtering, Content-based system and product affinity

Architect-ed A/B testing dashboard for easy monitoring performance of 3+ purposes: recommendation engine, analyze & compare the trends and the product journey

• Team management and delegation of work (model development and improvement in the existing models) among team members
• Machine Leaning Model Development

Deployment of 4 machine learning models: churn prediction, recommendation system, propensity to convert and multi-sim

Dealt with on-shore and 6 off-shore clients in Kenya, Afghanistan,Oman, Domino Republic, Ghana, Colombia.

Total deliveries of 14 models in the year 2019-2020

• Research and Development

Acceptance of research publication on recommendation system at Malaysia Conference by IEEE

Rectification of existing A/B testing dashboard and further its development in pyspark for client in Kenya

Involved in improving existing recommendation system to align it with the client's market purchase trends.

• Key Achievements

Onsite deployment of 3 models at Oman within 14 days, and cleared the delivery of ~40% of pending project revenue.

• Research and Development

Thorough literature review and presentation to IITb professor Dr. Animesh Kumar on recommendation system developed in-house.

• Machine Leaning Model Deployments

Accomplished the deployment of churn model at offshore clients in Ghana and Colombia

Succeeded in PoC deployment for offshore client in Austria

• Principal Investigator

Synthesis and characterization of novel sophorolipids and their therapeutic efficacy in vitro

• Research and Development

Assisted in developing R&D center

Handled an industrial scale project for the elimination of toxins, mainly aflatoxins, from different food commodities.