Sharad Pal

AP P2P $STP_Gap Reduction :

· $STP_Gap: Days difference between Scheduled Due Date and Payment Date (Dollar Weighted)

· Objective was to reduce $STP_Gap for Ariba P2P invoices to align with Ariba online baseline and improve requestor behavior

· As a solution, created a ‘Top 50 Requestor Dashboard’ which provides a monthly view of all the relevant information i.e. $STP_Gap trend, Invoice journey, Top defaulters etc.

· Created a new metric called Impact to identify the top defaulters among all the requestors and thus help in ranking them

· Business got a complete view of requestors performance and STP_Gap$ trend.

· Invoice journey helped business to identify the areas of improvements and process changes are happening.

· Dashboard helped regional leaders to pin point ‘pain areas’ easily.

· Significant improvement in STP_Gap$ for EMEA i.e. From 38 days to 13 days in last 4 months(Q4_M3-Q2_M1).

Automated Insight Generator/PD Root Cause Analysis :

· Objective is to help business derive insights from PD combined ledger data and provide quick action item which will help to reduce PD

· Created a Decomposition Tree Visual for current PD and PD Delta (QoQ) with all the significant drivers and top accounts against each RCA path with invoice details

AP Controls Dashboard

· Objective of this request is to identify the AP payments which are supposed to go through ARIBA but not going through ARIBA

· Created a PowerBI dashboard for Sundry and Non-Sundry payments which will help business to identify the top drivers this behavior and will enable better governance around it

· Created different logic and implemented into dashboard to identify wrong payments for Sundry and Non-Sundry payments

· Dashboard is in production.

FSR Forms comment analysis:

· Analyzed historical data of AP FSR forms of last 2 year and did topic modeling against Requester, Processor and approver comments

· Identified the variance between form names and actual issues. Business is utilizing this data to fill this gap.

AP Approval Threshold Analysis :

· Objective of this request was to come up with data driven threshold for AP invoices approval currently business follows the logic that any invoice which is >$20k needs two approval rest just one

· Analyzed the invoice amount and invoice volume data of 2 quarters and studied its distribution

· Leveraging this historical data created two set of approval matrix with 2 and 4 levels of approvals

·Its being reviewed by regional leads.

MPSR indirect variables impact analysis:

· MPSR- Minutes Per Service Request, is a key metric which helps business to decide the payout for OSPs

· Objective of this project is to derive the impact of all the metrics which are not directly going into the formulae of MPSR and have a significant impact on it, so business can control these metrics and thus keep the value of MPSR low.

· Used step wise elimination regression analysis to fix on the significant matrices by feeding the last two years of data in the model.

· Successfully overcame the challenge of associating the unit change of dependent variable to independent variables, which was the key ask from partner, by leveraging the ‘due-to’ analysis.

· Proposed the SPC control limits to further monitor the performance of indirect metrics and team has started implementing this solution.

· Worked on creating the automated process to scale this solution to different business regions and segments.

APOS Lead Generation:

· Objective of the project is to increase lead base and improve ranking of target customers

· Increased the lead base by ~25% by taking this process in house.

· Used LogR and 'Random Forest' to improve ranking of lead base since this model can assign a particular probability score to every lead and leads can be ranked in a better way

· With this better lead base and ranking business achieved highest RPL of $5

· Due to huge success of this pilot in North America process got accepted globally.

Support Assist Toaster Campaign/White Space Analysis

· Objective of this POC is to create a reporting solution/ Visualization for support assist toaster campaign by mapping this SA data to OMNEO and TD.

· White space is defined as the total out of warranty tags for which diagnostic test has failed in OMNEO and customer has not contacted dell.

· Created KPIs in OMNEO to get required data set for SA_Diags and SA_Alerts. Analyzed various OMNEO templates to get all the columns at one place and since OMNEO has a limitation of extracting only 1 million rows (with duplicates) per export it became a multiple export and a huge data crunching problem to get just one view.

· Created a new search template in OMNEO to get the required data for white space analysis and then exported and combined this data for further mapping it to TD data.

Process Digitization using ARIS

· Objective of this POC is to create the digital blueprint of different business processes in SDS domain by harnessing the power of BPMN2.0 in ARIS and later optimize those processes by leveraging the simulation feature of ARIS.

· Went through the training videos and other study material shared by business to help myself to get familiarized with ARIS and a tool and its capabilities.

· Analyzed a case study on ARIS implementation in ProDeploy for Client and scrutinized the approach that was used and thus doing so, came across the avenues for improvement of the analysis.

· Went through the formal training on ARIS connect and created few test models and validated them.

· Created a process map for APOS renewal process.

Proactive Escalation Management:

· Objective of this POC was to predict escalations from ongoing SRs so that APJ tech support team(RMs) can proactively contact the customer and resolve the issue and thus reduce the OPEX and improve CSAT

· Went through the RPN method used by business to predict the escalations and found that there is a lot of scope of improvement in terms of accuracy and efficiency

· Did the historical SR escalation analysis on one and half year data and considered different customer level and SR level Variables

· Analyzed the TTE (Time to escalate) and found that within 10 days of escalation creation 80% SRs may escalate, this analysis helped us in deciding the critical time limit to measure the efficiency of model.

· Based on this historical data, implemented several classification algorithms i.e. Random forest & LOGR, and predicted the escalations with Prob. Score with Model accuracy of 96%

· Achieved a huge increment in terms of Escalation prediction accuracy of 20% in comparison to RPN model.

· Did the live SRs prediction on a daily basis and achieved 66% accuracy of escalation prediction

· Provided optimal count of SRs i.e. sweet spot analysis for Proactive calling so that business could achieve maximal escalation prevention with minimal proactive calls

· Contributed in writing the White Paper for the process used in this POC. White paper got successfully submitted and is now available on dell site.

Service Opportunity Analysis:

· Objective of this POC is to compare the deal conversion of hardware sales when a service is attached or not and pitch this result to the Sales team to derive the future strategy

· Initial approach was to take Deal data from SFDC and map it back to order data from TD but due to lack of common link between TD and SFDC this approach was not fruitful. Also discussed this issue to the various teams across the floor but all discussion resulted in same. shared this feedback with the partner

· Shared the first cut of report on a sample data of FY’18 Q1 consisting of diverse services, their different permutations & combinations with their attach rates, revenue & margin

· Also suggested the business a what if analysis using the POWER BI dashboard in which we showcased them how revenue can be increased by pitching one combination of services over another for different LOBs

· Presented the report for extended time frame of one year (FY’17 Q4 to FY’18 Q3)

Early Warning System:

•Objective is to reduce the no of dispatches at commodity level.

•Built a SQL rule engine with inbuilt SPC algorithm to filter out the violations among all the dispatches

•Used trend analysis with SPC on the historical performance of each violation and reported it as a trigger to business

DSP Comments Analysis:

· Objective of this POC was to measure the field engineers’ compliance based on comments (‘Fault’ & ‘Repair’) added by them for EMEA region.

· Created a new metric ‘Compliance Percentage’ to measure the compliance of field engineers

· Compliance percentage is calculated by taking the ratio of Dispatches with compliant comment vs Total Dispatches (MDR 1)

· Reported to the business that current compliance rate is 4.3 % in EMEA and presented them different cuts of this report telling how compliance is varying by country, LOB & week