Cadet in the National Cadet Corps contingent in the institute, participated in the ‘Swachh Bharat Abhiyan’

Designed and developed a rule-based chatbot from scratch by implementing a hierarchical intent matching mechanism and provided service as an API, efficiently handling 85% of inbound messages in production (Python)

Automated the customer disenrollment process at the backend, saving nearly 10 hours per week for CSMs

Implemented, tested, and maintained SMS, IVR, and Email campaigns logic in PHP using the Symfony framework

Implemented a linear optimization algorithm to divide the USA into sales zones achieving optimal user response 
Developed a python tool to automate the visualization of the zones using folium, saving nearly 200 analyst hours
Generated an additional $2.16 Million in revenue by improving the response rate and the sale event size

Loylty Rewardz - Loyalty program management arm of Billdesk
Implemented XGBoost library and LIBSVM in python, reducing model runtime by 40% and improving 5X efficacy
Applied the Pareto/NBD model with Markov-Chain-Monte-Carlo simulation for parameter estimation to track customer buying patterns along with CLV and consequently filter out customers at the risk of attrition
Improved 1.5% efficacy by performing Bayesian A/B testing on reactivation campaigns for dormant customers
Collaborated with bank CXO’s to setup Tableau dashboard with over 40 insightful reports each for 25 banks 
Utilized Gestalt Pattern Matching to improve data quality by 5% compared to Levenshtein edit distance method

Bypassed computationally expensive GW electronic bandgap calculations using machine learning
Enhanced the existing feature set by inducing non-linearity in features and utilizing LASSO for feature selection 
Implemented 10-fold cross-validated kernel-based Gaussian Process Regression to improve accuracy by 15%
Publication: Machine-Learning -Assisted Accurate Band Gap Predictions of Functionalized MXene, 2018

Instichoose,​ ​ Co-founder​ ​ and​ ​ Chief​ ​ Product​ ​ Officer 

- Co-founded​ ​ a ​ ​ seed​ ​ funded​ ​ startup​​ ​ which​ ​ aimed​ ​ at​ ​ revolutionizing​ ​ the​ ​ problem​ ​ of​ ​ school​ ​ admissions
- Instichoose​ ​ provided​ ​ online​ ​ platform​ ​ for​ ​ advanced​ ​ school​ ​ search​ ​ and​ ​ send​ ​ online​ ​ applications​ ​ to​ ​ institutes
− Conceptualized​ ​ and​ ​ designed​ ​ the​​ ​ UI/UX​ ​ of​ ​ the​ ​ platform​ ​ to​ ​ optimize​ ​ user​ ​ flow​​ ​ and​ ​ build​ ​ intuitive​ ​ platform
− Onboarded​ ​ 15+​ ​ schools​ ​ for​ ​ beta​ ​ testing​​ ​ and​ ​ made​ ​ product​ ​ iterations​ ​ based​ ​ on​ ​ both​ ​ school​ ​ and​ ​ user​ ​ needs
− Built​ ​ a ​ ​ data​ ​ driven​ ​ school​ ​ rating​ ​ system,​ ​ further​ ​ used​ ​ it​ ​ for​ ​ school​ ​ recommendation​ ​ system​ ​ and​ ​ analytics
− Operational​ ​ in​ ​ 13​ ​ cities​ ​ with​ ​ over​ ​ 4000​ ​ institutes​ ​ listed​ ​ and​ ​ received​ ​ seed​ ​ commitment​ ​ of​ ​ INR​ ​ 25​ ​ Lacs
− Featured​ ​ on​ ​ front​ ​ pag

Software Engineer with an appetite for ML.

Rishabh Vaish | Peerlist Profile

Walmart sale in stormy weather

Wine Reviews

Applied the Pareto/NBD model with Markov-Chain-Monte-Carlo simulation for parameter estimation to track customer buying patterns along with CLV and consequently filter out customers at the risk of attrition

Attrition Model

-Accelerated​ ​ the​ ​ traditional​ ​ computational​ ​ bandgap​ ​ prediction​ ​ procedure​ ​ by​ ​ developing​ ​ alternate​ ​ quantitative​ ​ regressive​ ​ model
-Generated​ ​ nonlinear​ ​ feature​ ​ space​​ ​ and​ ​ implemented​ ​ cross​ ​

Machine​ ​ learning​ ​ augmented​ ​ computational​ ​ material​ ​ science​ ​

- Collaborated​ ​ with​ ​ Korea​ ​ Institute​ ​ of​ ​ Science​ ​ and​ ​ Technology​​ ​ to​ ​ screen​ ​ MXene​ ​ for​ ​ photocatalysis​ ​ by​ ​ creating​ ​ prediction​ ​ model
− Implemented​​ ​ feature​ ​ engineering​ ​ technique​ ​ to​ ​ ge

Photocatalyst​ ​ screening​ ​ using​ ​ machine​ ​ learning​ ​ ​

- Correlated​ ​ corruption​ ​ index​ ​ with​ ​ income​ ​ inequality​ ​ using​ ​ multivariate​ ​ regression​​ ​ of​ ​ corruption​ ​ with​ ​ independent​ ​ variables
− Studied​ ​ the​ ​ negative​ ​ effect​ ​ of​ ​ high​ ​ democracy​ ​ levels​

Vicious​ ​ circle​ ​ of​ ​ corruption​ ​ and​ ​ inequality​ ​

-Investigated the cause of fast growth rate of thailand considering both immediate factors and underlying social and political factors using the datasets provided by world bank and bank of thailand

-Concluded the effects of dynamic entrepr

Economic Growth in Thailand: Macroeconomics

-Investigated the relationship between economic growth rate and other economic factors savings rate, productivity, investments,population, GDP applying solow growth model on the data of 190 countries

-Recommended policies tax rates, invest

Analyses of factors of economic growth

-Integrated customer behavioral variables(RFM), demographic variables, and transaction database to create a method of mining changes in customer behavior using apriori algorithm and association rule mining

-Used two measures of similarity 

Data Mining in customer behavior changes

-Analyzed the dataset of a telecom company, Cell2Cell, to make a predictive model that gave churning probability and implemented data cleaning, data pre-processing and attribute reduction techniques Correlation Matrix, Similarity - Dissimil

Rishabh Vaish

Software Engineer with an appetite for ML.

Walmart sale in stormy weather

Wine Reviews

Attrition Model

Machine​ ​ learning​ ​ augmented​ ​ computational​ ​ material​ ​ science​ ​

Machine learning augmented computational material science