際際滷

際際滷Share a Scribd company logo

Resume_Mohammed_jawed_khan

Jawed Khan
Jawed Khan

Mohammed Jawed Khan is an Indian national who currently works as a Data Scientist in Saudi Arabia. He has over 15 years of experience in business analytics, data science, and machine learning. He holds a Master's degree in Business Analytics and has received several certifications in predictive modeling, Six Sigma, and SAS programming. Khan has worked on a variety of analytics projects for companies in sectors like automotive financing, insurance, and energy. His responsibilities have included developing predictive models, forecasts, marketing campaigns, and tools to optimize business processes.

1 of 5
Download to read offline
1/5 Resume Personal Details Name Mohammed Jawed Khan Email mohjkhan@indiana.edu Mobile +966-568121688 Nationality Indian Marital Status Married Date of Birth 03 Oct 1975 Role: Data Scientist Qualifications:- Graduation: Master of Science, Business Analytics (2015) Kelley School of Business, Indiana University, USA. Under Graduation: Bachelor of Technology, I.I.T New Delhi (2001) International Languages:- i. English (TOEFL-293/300 and 5.5/6 essay) ii. German (DSH Universit辰t Stuttgart-Bestanden). Certification:- i. Graduate Certificate in Business Analytics (2014) IIM-Lucknow. ii. Certified Predictive Modeller SAS Enterprise Miner 7 from SAS Institute iii. Six- Sigma Master Black Belt Certification from Indian Statistical Institute. Business Analytics Tools:- Oracle Crystal Ball, Advance Excel (@RISK and the Decision Tools Suite, Power Pivot), POM-QM. Python, Matlab, SAS Enterprise Miner, JMP, SPSS. BI-Tools: QlikView, Tableau, Excel Visualization Programming:- Base SAS, Java/ J2EE, Python, R, Visual Basic, MS/Unix Script, PL/SQL
2/5 Professional Experiences:- 1. Operation Management (ALJF) (Jan12- current) Role: Data Analyst; Location: KSA About Company: ALJUF is the market leader in providing Auto finance to its customers whether fleet or consumer finance. It facilitates sales of new and used Toyota & Lexus vehicles through its leasing and instalment portfolios. It has a network over 230 branches spread over the country and its current receivables is around 17 billion SAR. My role in the company has been initially to initiate Risk Modelling solutions for various departments using COSO framework, develop resource and process to mature ALJ in Analytics based company, and finally lead the road map to the ambitious and high invested endeavour to integrate Enterprise Analytics. i. Maximize collectors performance. Challenge: ALJUF is reducing number of collectors which will reduce available collection efforts. This may directly impact the overdue of the company. The task was to optimize available collection efforts (call, SMS, legal action) with the limited number of collectors without adversely affecting the overdue of the company or alienating profitable customers. Collection effort was customized according to identified risk groups. Based on customer characteristics, a statistical model based on survival analytics is developed, to identify by when a particular guest should have paid their overdue. If payment was not received by this time, the system triggered Agents to apply or intensify collection efforts. Different collections efforts have different cost, and designing appropriate effort will optimize collection operation cost. Specific risk profile guest receive customized messages at appropriate time, providing them time to act before the next level collection effort is employed. Guests are called only when payment is not received until the model suggested time. This will reduce the number of Guests to be called by the collectors thus reducing collectors call effort. Guests whose payment time is suggested by the model beyond the current month should be proactively contacted from the beginning of the month thus avoiding their delinquency. Collectors receive their specific list to call/SMS which facilitates them to optimize their work. Customer feedback using text mining was used to update customer records and sentiments/follow ups. ii. Target Forecast System: - ALJL requires forecasts of sales and collection projects monthly targets at each hierarchy of branch staff, supervisor, branch manager, AGM and Director based on historical patterns using statistical methods. This is used to evaluate performance of branch staff and evaluate commission of staff of
3/5 profit centres. The historical time series sales and collection monthly data of length 10 years was collected from FS-System. The seasonal effect and periodicity was identified and filtered using adjusted moving average method. The remaining effects were those of cyclic components which was modelled using Fourier analysis and distinct frequencies were identified. Finally diagnostic checks were performed on residual (irregular) as free from autocorrelation or partial autocorrelations till 30 lags through Box-pierce statistics. iii. Marketing Campaign: Led the development of Analytic List for targeted solicitation of potential repeat customer. The ensemble model prioritise potential customer based on recommended probability from decision tree, neural network, gradient boosting, logistic regression. The recommending also suggested the preferred product and the time of solicitation based on Survival Analytics. The model suggested the communication channel as well as the time to broadcast the promotion for specific customer. Using text analytics of campaign feedback, campaign analysis success was measured. iv. Application Score Card: - I design and development of Customer Score Card (customer credit worthiness assessment) based on training a predictive model using Statistical methods with existing customer based on their payment behaviour and customer demographic profile. This model predicts score evaluated for each application. The trained model through regression model, decision tree and neural net provided prediction on test data with reasonable accuracy. Thus a predictive engine for new customers provided a score which became a basis of considering them for new lease application. v. Bad Debt provisioning criteria: - The Company has every month outstanding receivables on which bad debt provision is to be ascertained in an optimal way using Flow method Approach. Based on historical collection pattern, a series of segments were created based on activity and months overdue applied at contract level thus achieving lower bad debt provisioning to Basel II capital framework. 2. IBM Research (Oct06- Jan12) Role: Business Analyst Location: Stuttgart, Kassel (Germany) i. Daimler AG: IBM Daimler Chrysler (IVK) was developed to solve Constructive Problems during Configuration as Model Construction for customization of design and configuration of Lastwagen. Constraint solver Engine was IBM product and we utilized to compute minimum conflict goals while solving interactive system constraints during dynamic formulation of Daimler vehicles. The challenge was to translate vehicle configuration constrains into proportional logic to model dependencies and finite domain constraints to represent conflicts. We also customised the Constraint Solvers minimal conflict solver engine to pick solution from sub optimal space to provide for consideration on time and memory optimisation, even though constraint suspension was accommodated
4/5 by tolerable threshold. This was quite an innovation and stimulating team from cross cultural team from IBM Germany, India and Israel and we could implement state of art solution. ii. Volkswagen: Volkswagen AG Ersatzteile 2000 is an IT-logistic parts ordering system forms the backbone for the supply of spare parts for the VW and Audi dealerships Kassel warehouse. ET2000 had capacity issues and poor responsiveness due to limitation of their base EOQ model in handling some of the slow moving parts which always had to be backordered at a cost, hence enhancement was required. Using Demand forecasts of each slow moving Auto parts based on historical data on demand and model variability in lead time and applying Quantity discount models, a stochastic model to control inventory was developed to generate reorder quantity and reorder schedule to optimize inventory at each Tier of integrated distribution system. To avoid stock outages, safety stock amount was derived from the stochastic model. 3. Morgan Stanley (Apr04 2004 till Oct06) Role: Associate; Location: Geneva (Switzerland). Projects: MSCI BARRA Index creation MSCI provides global equity indices, which, over the last 30+ years, have become the most widely used international equity benchmarks by institutional investors. Barra is the market leader in delivering innovative, financial risk management solutions worldwide. I was appointed as Associate and my work comprised of Modelling & Analysis of MSCI indices creating market segmentation of portfolio grouping for advising investors. Capital Investment for MSCI Barra Index at Geneva which resulted in optimal portfolio creation based on Risk and Return Period Profiles. 4. Project Associate, IKE, Germany (Feb02 Mar06). Role: Associate; Location: Stuttgart, Germany Projects: ECA This project was developed by Instit端t f端r Kernenergie Energiesystem (IKE) to create simulation model of Energy usage in school building. A generic data model was used to define building components and associating energy loss parameters. The model was then subjected to different schedule profiles and subjected to varying Heating, Ventilation and Air Conditioning (HVAC) conditions. This resulting energy usage was fed to the Energy Concept Advisor (ECA) tool which was developed to give retrofitting advice for building energy optimization. 5. AIIMS, New Delhi (Jan01 Jan02) Role: Associate; Location: New Delhi, India Projects: Department of Science and Technology, Government of India The project was to create a disability correction tool through presenting acoustically modified learning programs for dyslexic children.
5/5 Publication: Policy paper on Energy Concept Advisor. http://www.annex36.de/eca/de/06util/pdf/A36SubtaskC_Appendix_ECA.pdf Training/Internship May00 - Aug00 Research Associate, Lehrstuhl f端r Mustererkennung und Bildverarbeitung, Instit端t f端r Informatics; Freiburg Universit辰t. Project I: Development of correction parameter table for a scanner which scanned a colour palette of combination of 64 colours and showed discoloration. With statistical methods, the colour table could map to its original colour combination from RGB. Project II: Robot vision program which identified the corners of cubes using shortest path technique where curvature of line plot showed rapid change is curvature. Recommendation: 1. Prof Dr Fritz Schmidt (IKE, Stuttgart University) - Pensioner 2. Prof Dr Ash Soni (KSB, Indiana University) 3. Prof Dr Venkataramanan (KSB, Indiana University) 4. Prof Dr Frank Acito (KSB, Indiana University) 5. Prof Dr Yogesh Agarwal (Indian Institute of Management, Lucknow) 6. Prof Dr Gaurav Garg (Indian Institute of Management, Lucknow) 7. Prof Dr Amit Agrahari (Indian Institute of Management, Lucknow) 8. Prof Dr R K Srivastava (Indian Institute of Management, Lucknow) 9. Prof Dr Veena Kalra (All India Institute of Medical Science, Delhi) 10.Prof Dr Prem Kalra (Indian Institute of Technology, Delhi)

More Related Content

Resume_Mohammed_jawed_khan

  • 1. 1/5 Resume Personal Details Name Mohammed Jawed Khan Email mohjkhan@indiana.edu Mobile +966-568121688 Nationality Indian Marital Status Married Date of Birth 03 Oct 1975 Role: Data Scientist Qualifications:- Graduation: Master of Science, Business Analytics (2015) Kelley School of Business, Indiana University, USA. Under Graduation: Bachelor of Technology, I.I.T New Delhi (2001) International Languages:- i. English (TOEFL-293/300 and 5.5/6 essay) ii. German (DSH Universit辰t Stuttgart-Bestanden). Certification:- i. Graduate Certificate in Business Analytics (2014) IIM-Lucknow. ii. Certified Predictive Modeller SAS Enterprise Miner 7 from SAS Institute iii. Six- Sigma Master Black Belt Certification from Indian Statistical Institute. Business Analytics Tools:- Oracle Crystal Ball, Advance Excel (@RISK and the Decision Tools Suite, Power Pivot), POM-QM. Python, Matlab, SAS Enterprise Miner, JMP, SPSS. BI-Tools: QlikView, Tableau, Excel Visualization Programming:- Base SAS, Java/ J2EE, Python, R, Visual Basic, MS/Unix Script, PL/SQL
  • 2. 2/5 Professional Experiences:- 1. Operation Management (ALJF) (Jan12- current) Role: Data Analyst; Location: KSA About Company: ALJUF is the market leader in providing Auto finance to its customers whether fleet or consumer finance. It facilitates sales of new and used Toyota & Lexus vehicles through its leasing and instalment portfolios. It has a network over 230 branches spread over the country and its current receivables is around 17 billion SAR. My role in the company has been initially to initiate Risk Modelling solutions for various departments using COSO framework, develop resource and process to mature ALJ in Analytics based company, and finally lead the road map to the ambitious and high invested endeavour to integrate Enterprise Analytics. i. Maximize collectors performance. Challenge: ALJUF is reducing number of collectors which will reduce available collection efforts. This may directly impact the overdue of the company. The task was to optimize available collection efforts (call, SMS, legal action) with the limited number of collectors without adversely affecting the overdue of the company or alienating profitable customers. Collection effort was customized according to identified risk groups. Based on customer characteristics, a statistical model based on survival analytics is developed, to identify by when a particular guest should have paid their overdue. If payment was not received by this time, the system triggered Agents to apply or intensify collection efforts. Different collections efforts have different cost, and designing appropriate effort will optimize collection operation cost. Specific risk profile guest receive customized messages at appropriate time, providing them time to act before the next level collection effort is employed. Guests are called only when payment is not received until the model suggested time. This will reduce the number of Guests to be called by the collectors thus reducing collectors call effort. Guests whose payment time is suggested by the model beyond the current month should be proactively contacted from the beginning of the month thus avoiding their delinquency. Collectors receive their specific list to call/SMS which facilitates them to optimize their work. Customer feedback using text mining was used to update customer records and sentiments/follow ups. ii. Target Forecast System: - ALJL requires forecasts of sales and collection projects monthly targets at each hierarchy of branch staff, supervisor, branch manager, AGM and Director based on historical patterns using statistical methods. This is used to evaluate performance of branch staff and evaluate commission of staff of
  • 3. 3/5 profit centres. The historical time series sales and collection monthly data of length 10 years was collected from FS-System. The seasonal effect and periodicity was identified and filtered using adjusted moving average method. The remaining effects were those of cyclic components which was modelled using Fourier analysis and distinct frequencies were identified. Finally diagnostic checks were performed on residual (irregular) as free from autocorrelation or partial autocorrelations till 30 lags through Box-pierce statistics. iii. Marketing Campaign: Led the development of Analytic List for targeted solicitation of potential repeat customer. The ensemble model prioritise potential customer based on recommended probability from decision tree, neural network, gradient boosting, logistic regression. The recommending also suggested the preferred product and the time of solicitation based on Survival Analytics. The model suggested the communication channel as well as the time to broadcast the promotion for specific customer. Using text analytics of campaign feedback, campaign analysis success was measured. iv. Application Score Card: - I design and development of Customer Score Card (customer credit worthiness assessment) based on training a predictive model using Statistical methods with existing customer based on their payment behaviour and customer demographic profile. This model predicts score evaluated for each application. The trained model through regression model, decision tree and neural net provided prediction on test data with reasonable accuracy. Thus a predictive engine for new customers provided a score which became a basis of considering them for new lease application. v. Bad Debt provisioning criteria: - The Company has every month outstanding receivables on which bad debt provision is to be ascertained in an optimal way using Flow method Approach. Based on historical collection pattern, a series of segments were created based on activity and months overdue applied at contract level thus achieving lower bad debt provisioning to Basel II capital framework. 2. IBM Research (Oct06- Jan12) Role: Business Analyst Location: Stuttgart, Kassel (Germany) i. Daimler AG: IBM Daimler Chrysler (IVK) was developed to solve Constructive Problems during Configuration as Model Construction for customization of design and configuration of Lastwagen. Constraint solver Engine was IBM product and we utilized to compute minimum conflict goals while solving interactive system constraints during dynamic formulation of Daimler vehicles. The challenge was to translate vehicle configuration constrains into proportional logic to model dependencies and finite domain constraints to represent conflicts. We also customised the Constraint Solvers minimal conflict solver engine to pick solution from sub optimal space to provide for consideration on time and memory optimisation, even though constraint suspension was accommodated
  • 4. 4/5 by tolerable threshold. This was quite an innovation and stimulating team from cross cultural team from IBM Germany, India and Israel and we could implement state of art solution. ii. Volkswagen: Volkswagen AG Ersatzteile 2000 is an IT-logistic parts ordering system forms the backbone for the supply of spare parts for the VW and Audi dealerships Kassel warehouse. ET2000 had capacity issues and poor responsiveness due to limitation of their base EOQ model in handling some of the slow moving parts which always had to be backordered at a cost, hence enhancement was required. Using Demand forecasts of each slow moving Auto parts based on historical data on demand and model variability in lead time and applying Quantity discount models, a stochastic model to control inventory was developed to generate reorder quantity and reorder schedule to optimize inventory at each Tier of integrated distribution system. To avoid stock outages, safety stock amount was derived from the stochastic model. 3. Morgan Stanley (Apr04 2004 till Oct06) Role: Associate; Location: Geneva (Switzerland). Projects: MSCI BARRA Index creation MSCI provides global equity indices, which, over the last 30+ years, have become the most widely used international equity benchmarks by institutional investors. Barra is the market leader in delivering innovative, financial risk management solutions worldwide. I was appointed as Associate and my work comprised of Modelling & Analysis of MSCI indices creating market segmentation of portfolio grouping for advising investors. Capital Investment for MSCI Barra Index at Geneva which resulted in optimal portfolio creation based on Risk and Return Period Profiles. 4. Project Associate, IKE, Germany (Feb02 Mar06). Role: Associate; Location: Stuttgart, Germany Projects: ECA This project was developed by Instit端t f端r Kernenergie Energiesystem (IKE) to create simulation model of Energy usage in school building. A generic data model was used to define building components and associating energy loss parameters. The model was then subjected to different schedule profiles and subjected to varying Heating, Ventilation and Air Conditioning (HVAC) conditions. This resulting energy usage was fed to the Energy Concept Advisor (ECA) tool which was developed to give retrofitting advice for building energy optimization. 5. AIIMS, New Delhi (Jan01 Jan02) Role: Associate; Location: New Delhi, India Projects: Department of Science and Technology, Government of India The project was to create a disability correction tool through presenting acoustically modified learning programs for dyslexic children.
  • 5. 5/5 Publication: Policy paper on Energy Concept Advisor. http://www.annex36.de/eca/de/06util/pdf/A36SubtaskC_Appendix_ECA.pdf Training/Internship May00 - Aug00 Research Associate, Lehrstuhl f端r Mustererkennung und Bildverarbeitung, Instit端t f端r Informatics; Freiburg Universit辰t. Project I: Development of correction parameter table for a scanner which scanned a colour palette of combination of 64 colours and showed discoloration. With statistical methods, the colour table could map to its original colour combination from RGB. Project II: Robot vision program which identified the corners of cubes using shortest path technique where curvature of line plot showed rapid change is curvature. Recommendation: 1. Prof Dr Fritz Schmidt (IKE, Stuttgart University) - Pensioner 2. Prof Dr Ash Soni (KSB, Indiana University) 3. Prof Dr Venkataramanan (KSB, Indiana University) 4. Prof Dr Frank Acito (KSB, Indiana University) 5. Prof Dr Yogesh Agarwal (Indian Institute of Management, Lucknow) 6. Prof Dr Gaurav Garg (Indian Institute of Management, Lucknow) 7. Prof Dr Amit Agrahari (Indian Institute of Management, Lucknow) 8. Prof Dr R K Srivastava (Indian Institute of Management, Lucknow) 9. Prof Dr Veena Kalra (All India Institute of Medical Science, Delhi) 10.Prof Dr Prem Kalra (Indian Institute of Technology, Delhi)