1 of 39

2 of 39

Cycle Time Reduction for POs Processing in Commercial Data Warehouse

Project ID	turbostrive-ISAC-2014-4758
Green Belt	G Gayatri Soumya (86019)
Black Belt	Rajaram G (RG00327136)
Master Black Belt	Nanjunda Swamy (127049)
RM	Sarada Rao (27446)
Sponsor	Neal Jacob (952)

3 of 39

Project Summary

Business

Problem

Statistical

Problem

Statistical�Solution

Business

Solution

Commercial Data Warehouse is Integrated with multiple ERP sources and Cycle time in processing purchase orders is High due to manual efforts which causes delay in Purchase Order Amounts reporting. Thus impacting calculation of incentives for GE sales executives after every quarter .
Poor Data Quality of customer details causes recurring load cycle.
Performance of data loading window is low.

Purchase Orders from multiple ERP systems are received during every quarter end from Power and Water ,Oil and Gas Businesses and processing them manually takes 6 weeks .
Reprocessing rejections in POs, Corrections in customer details delays reporting for another 2 weeks.
End to end process requires all team members engaged for 8 weeks.

- Per Feed > 24hrs (41 Feeds per Quarter ) Multiple iterations do be done until the final feed is with proper customer details. ( On an average 8 Weeks to complete the E2E cycle )

End to end process completes in 4 weeks.
Complete Process requires only 1 team member to monitor loads (if failures)

Manual Efforts –Automated
Data Quality Improved
Tuned Production loading window time

4 of 39

Project Summary

Project Name : Commercial Reporting Project ID : 13001040

GEPS25 Orders & Sales Metric project will focus on supplying the Digital Cockpit with the disparate Orders & Sales data from across GEPS(General Electric Power Systems). We report Orders data like Order amount, customer details for O&G, P&W, EM businesses . The purpose of the application is to provide an intensive decision support system for the GEPS management to identify top customers and derive incentives/variable compensation for GE sales executives based on Order amounts reported .

We get Orders source feeds from around 41 source systems and details need to be reported after loading into ODS. ODS is Operational Data Store which was build on various Business/GAP rules to report accurate Information.Each source being an individual ERP, raw data is taken and being manually formatted for 50% percent of source feeds. Due to the manual efforts, it takes around 6 weeks of time to load and another 2 weeks to correct the customer details if any. If data has more rejections ,then this may delay in reporting . This would impact Business since Salaries of GE employees are affected. Manual effort for rejections/corrections, Data loading window cycle are the two pain areas.

Below are the pain areas which should be taken into consideration and handled –

Manual Processing of ERP Orders ,Reprocessing of Rejections/corrections.
High Cycle time for Data loading window in QA and Production environments.

Executive Summary

5 of 39

DEFINE

6 of 39

Voice of Stakeholders

Associate

What they said ? (Voice of Associate)	What they meant ?
What they said ? (Voice of Associate)	What is the Need ?	When is the Need felt ?	Where is the Need felt ?	Why is the Need felt ?	How is the process now ?
Huge amount of manual efforts are involved in processing purchase orders since all businesses are not able to send the Orders in a standard format.	To process the Purchase Orders within 24 hrs. and report and to avoid manual efforts	When feed is bulk and manual efforts consume long time due to more number of rejections.	Need felt while processing the Purchase Orders and validating the rejections.	To provide the synchronized Purchase Orders DWH report to users in business hours	Currently Purchase Orders are processed Manually and tested in QA environment for validation of rejections because of which each feed takes >24 hrs. to process.
Cycle time of entire PO processing is high since Purchase Orders are being sent by multiple business from different time zones and location.	Purchase Orders to be reported on the same day as and when they are received irrespective of time of arrival from Business	When users from multiple businesses send feed during different time zones and when team not available.	While Production loading window consume more time.	To make the DWH Sales/Orders reports available to Business for calculation of VC for sales executions on quarterly basis.	Currently Cycle time and Loading window is high since feeds come from different Businesses and in different time zones.

What they said ? (Voice of Investor)	What they meant ?
What they said ? (Voice of Investor)	What is the Need ?	When is the Need felt ?	Where is the Need felt ?	Why is the Need felt ?	How is the process now ?
The DWH should be enough robust and transparent for functional users to handle the customer needs and set a way to handle multiple businesses	Improve the goodwill of the DWH to business client	When customer is having dissatisfaction and asking for an better DWH	Report Availability for Finance team during calculation of salary (VC) for sales executives	To provide better solution to customer	Its good but not the best

Investor

7 of 39

Voice of Stakeholders

What they said ? (Voice of Customer)	What they meant ?
What they said ? (Voice of Customer)	What is the Need ?	When is the Need felt ?	Where is the Need felt ?	Why is the Need felt ?	How is the process now ?
Need a transparent tool to process the PO’s at any point of time and reports should be delivered on the same day.	For calculation of incentives in salaries of sales executives	When Purchase Orders are not received before final payroll cut off date.	To retrieve the DWH reports with synchronized data for all businesses within same day of PO arrival	Incentives getting impacted and users not able to view top customers information	Process not capable to deliver all business PO’s to users within required time period while accessing feeds from different time zones.(Business- Power & Water, Oil & Gas)

Customer (General Electric Inc.)

Society

What they said ? (Voice of Society)	What they meant ?
What they said ? (Voice of Society)	What is the Need ?	When is the Need felt ?	Where is the Need felt ?	Why is the Need felt ?	How is the process now ?
NA	NA	NA	NA	NA	NA

8 of 39

Voice of Stakeholders

S No	Voice of Stakeholders	Process / Service Attribute (refers to the AICS voice)	Need Statement �(From A-I-C-S Perspective)	Delight Parameter (Project y)
1	Cycle time of processing purchase Orders is high which is causing delay in calculation of incentives and effect salaries of GE sales executives. disturbs the budget allocation if incentives are carried forwarded to next month due to delay.	Delay in delivering Purchase Orders Data Warehouse (Reports) during quarterly ETL Load	Availability of Orders information for calculation of salaries	Reduction of cycle time of PO processing for On time calculation of salaries /incentives

Translation of Voice of Stakeholders

9 of 39

Project Charter

1.Commercial Data Warehouse is an integrated system (Operation Data Store) with multiple ERP sources which captures transactional level Purchase Orders of GE Power & Water, Oil & Gas Businesses. Cycle time in processing these purchase orders is High due to manual efforts which causes delay in Order Amounts reporting. Thus impacting calculation of Quarterly incentives for GE sales executives. Impact due to delay is high since GE Financial Budget allocation gets disturbed if incentives are carry forwarded to next month's salaries.
2.Poor Data Quality of customer details causes recurring load cycle. Thus inaccurate, varying Top Customer Reports are available for Business Users at GE headquarters level. Users needed Reports with good data quality.
3.Performance of data loading window is low.

Hence, the project needs to be initiated to reduce the Cycle time of PO processing and Reporting to business users and thus to provide a better solution to business.

Business Case

10 of 39

Project Charter

End to End load cycle for Purchase Orders reporting takes 8 weeks ,thus impacting incentive calculations.

Objective is to reduce cycle time to 50%. (8 weeks to 4 weeks)

Problem Statement

Goal Statement

11 of 39

Project Charter

Team Members Names	Roles & Responsibilities
G Gayatri Soumya	GB Trained team member.
Ankita Bhatia	Dev Team member
Dayananda Una	Dev Team member
Shanker Ravi	Production Support Team member

Key milestones/Team Roles

12 of 39

Macro Level Process Map

Supplier

Input

Process

Output

Customer

I

O

S

P

C

Unprocessed Purchase Orders from 20 Sub Businesses of GE P&W and O&G Businesses

Transactional Orders from each Sub Business ERP system.

Processing and Loading Purchase Orders into Operational Store(ODS)

Purchase Orders are updated in DWH and Reports are available for end users.

General Electric Inc.

Start

PO’s are manually processed and loaded in Quality Analysis environment. Then loaded in Production environment

Data loads into Staging Area

General Accounting Procedures/Business rules are applied and loaded into Commercial DWH after validation

Reports are refreshed with Purchase Orders details in DWH and sent to Functional team for Incentives/VC calculations

Stop

SIPOC Diagram

13 of 39

MEASURE

14 of 39

Data Collection Mechanism

41

Daily ( Quarter end activity)

Number of Data Points

Frequency of Data Points

15 of 39

Analysis Method

TechM Internal

Continuous

Type of Data

1.53

Baseline ST Sigma Level

16 of 39

Data Analysis

TechM Internal

DPMO Calculation

No of Defects : 20

No. of units processed: 41

No of OFE: 1


Defects per Million Opportunities	=	Number of defects found at any check point			X	1,000,000
		Number of Units processed	X	Number of oppurtunities of Error	X
		Number of Units processed	X	Number of oppurtunities of Error

Hence ,�� DPMO = 20 1,000,000 = 487805

----------- * � 41* 1

Using the Sigma tables, we can conclude that the Sigma value is 1.53

Defect : Data Feeds breaching SLA of 24hrs ��

17 of 39

ANALYZE

18 of 39

As-Is Value Stream Map

Analyze

Step 4

PO’s

Requested

Customer

5Min

Customer

PROCESSING

TIME

.

Manual

Processing

of PO

Identify

Raw feed

Screen

Door/QA

Process

Ticket

Assigned

Ticket

Logged

For Prod

QA

Testing

User

Verifies

POs

Provides

Approval

W

150Min

240Min

20Min

330Min

30Min

5Min

10Min

15Min

WAIT

TIME

15Min

30 min

15min

5 min

10 min

POs

Processed

Reports

Delivered

Manually

480Min

15Min

10 min

15 min

W

Non value added activity

Value added activity

Value enabling activity

Team Member

Support Member

Functional User

Datawarehouse Load

19 of 39

Process Value Link Example

Analyze

Step 4

20 of 39

Analyze

Waste Category Identification

Explain the percentage of VA , VE and NVA as in the current process using a pie chart

(This is not a continuation of example from previous slide)

VA:VE:NVA= 2:34:64

Step 4

VA(2%)

VE(34%)

NVA(64%)

21 of 39

Analyze

Waste Category Identification

VE-Value Enabler: Identifying Screen door rejections

T-Transportation : Manually Ftp feed to UNIX server, KM Portal.

All feeds are manually uploaded by customer ,downloaded from the library by team and tested files are manually uploaded in server and reports are uploaded in library

M- Motion: Logging tickets :-

Tickets logged, assigned, approved manually.

W- Waiting: Wastage of time:-

Waiting time to run the process , tickets to be assigned, executed in production, manual report extraction

O- Over Processing: Rework

Purchase orders are tested in quality analysis environment due to Rejections in source feeds and final numbers are validate by team and functional users before proceeding to production

D- Defects: Rejections in Purchase Orders

Purchase orders are reprocessing due to Rejections in source feeds. This indicates poor data quality

Step 4

22 of 39

Cause and Effect diagram

High Cycle Time for

PO Processing

Tools/System

Environment

Process

People

Ask “Why?”

Fishbone Diagram

Delay in Arrival of POs from Power & Water

And Oil & Gas Businesses

Huge Volume of Data Update at source

Transactional Order get rejected and reprocessed again

Data is sent by multiple businesses – Water, Aero, Gas Engines, Turbo Machinery etc.

Each ERP feed is generated from

different

applications

Due to GAP rules (General Accounting procedures)as per business requirement

Load is taking long time

Low throughput

Load Failure due to Source feed

Recurring load due to rejections

Temp space Full/ Lack of temp space

Data Issue

Database Issue

Database Table Locks

Incorrect Source Data

Load Failure based on target

Dependency on Multiple teams

Root Cause Analysis of any data mismatch

Manual Effort

ERP Source Teams Dependency

Support and Admin Team Dependency

Ticket logged to Support Team. They Work as per SLA

QA Load is to validate

rejections in data

Inaccurate/invalid data

Prod load wont contain rejected PO’s

Rejections if present need to be deleted and reloaded

Multiple Environments(QA/Prod)

Planned/Unplanned Outages

maintenance activities /Network Issues

Planned DB/ETL outages / Unable to connect servers

Manual Process consume more

Time in processing huge volume of Purchase Orders at one go.

Current process consumes manual efforts in converting source feed into standard format

Cause Level 1

Cause Level 2

Cause Level 3

Cause Level 4

Less feasibility by ERPs to generate PO information in standard format

Lack of Communication & coordination

23 of 39

Priority Matrix

Major Cause 1	Cause 1	Cause 2	Cause 3	Cause 4	Team member1	Team member2	Team member3	Team member4	Total
Tools / Systems	Load taking long time.	Low Throughput	Temp space Full/ Lack of temp space		2	3	2	3	10
	Load failure due to source	Data Issue	Incorrect Source Data		2	3	1	1	7
	Load Failure Based on Target	Database Issue	Database table Locks		1	2	1	1	5
Major Cause 2
Process	Delay in Arrival of POs from Power & Water And Oil & Gas Businesses	Each ERP feed is generated from different applications	Less feasibility by ERPs to generate PO information in standard format	Current process consumes manual efforts in converting source feed into standard format	5	5	4	4	18
	Transactional Order get rejected and reprocessed again	Due to GAP rules (General Accounting procedures)as per business requirement			3	3	2	2	10
	Huge Volume of Data Update at source	Data is updated for multiple businesses – Water, Aero, Gas Engines, Turbo Machinery etc.	Manual Process consume more Time in processing huge volume of Purchase Orders at one go.		5	5	4	4	18

Note: Higher the rating, Higher the impact on High Cycle time of PO Processing

24 of 39

Priority Matrix

Major Cause 3	Cause 1	Cause 2	Cause 3	Cause 4	Team member1	Team member2	Team member3	Team member4	Total
Environment	Multiple Environments(QA/Prod)	QA Load is to validate rejections in data	Inaccurate/invalid data		1	1	1	1	4
	Multiple Environments(QA/Prod)	Prod load wont contain rejected PO’s	Rejections if present need to be deleted and reloaded		1	2	1	1	5
	Planned Outages /Unplanned Outages	Planned DB/ETL /Unable to connect servers	Network Issues		2	2	2	2	8
Major Cause 4
People	Recurring load due to rejections	Root Cause Analysis of any failure	Manual Effort		2	3	2	3	10
	Dependency on Multiple teams	Support and Admin Team Dependency	Ticket logged to Support Team. They Work as per SLA		2	3	2	3	10
	Dependency on Multiple teams	ERP Source Teams Dependency	Lack of Communication & coordination		2	3	2	3	10

Note: Higher the rating, Higher the impact on Outage/downtime

25 of 39

Cause and Effect diagram(Vital ‘X’s identified)

High Cycle Time for

PO Processing

Tools/System

Environment

Process

People

Ask “Why?”

Fishbone Diagram

Delay in Arrival of POs from Power & Water

And Oil & Gas Businesses

Huge Volume of Data Update at source

Transactional Order get rejected and reprocessed again

Data is sent by multiple businesses – Water, Aero, Gas Engines, Turbo Machinery etc.

Each ERP feed is generated from

different

applications

Due to GAP rules (General Accounting procedures)as per business requirement

Load is taking long time

Low throughput

Load Failure due to Source feed

Recurring load due to rejections

Temp space Full/ Lack of temp space

Data Issue

Database Issue

Database Table Locks

Incorrect Source Data

Load Failure based on target

Dependency on Multiple teams

Root Cause Analysis of any data mismatch

Manual Effort

ERP Source Teams Dependency

Support and Admin Team Dependency

Ticket logged to Support Team. They Work as per SLA

QA Load is to validate rejections in data

Inaccurate/invalid data

Prod load wont contain rejected PO’s

Rejections if present need to be deleted and reloaded

Multiple Environments(QA/Prod)

Planned/Unplanned Outages

Planned DB/ETL outages / Unable to connect servers

Manual Process consume more

Time in processing huge volume of Purchase Orders at one go.

Current process consumes manual efforts in converting source feed into standard format

Cause Level 1

Cause Level 2

Cause Level 3

Cause Level 4

Less feasibility by ERPs to generate PO information in standard format

maintenance activities /Network Issues

Lack of Communication & coordination

26 of 39

Priority Matrix

Potential Root Causes: Justification –

We are considering rating 10 and above for the analyzing of Potential root causes.

Major Cause	Final Root Causes	Total Rating	Vital X	Justifications
Tools / Systems	Temp space Full/ Lack of temp space	10	Y
Process	Current process consumes manual efforts in converting source feed into standard format	18	Y
	Due to GAP rules (General Accounting procedures)as per business requirement	10	N	We don’t have any control over business requirements, hence not being considered for improvement, currently
	Manual Process consume more time in processing huge volume of Purchase Orders at one go.	18	Y
People	Manual Effort for Failure recovery	10	N	We don’t have any control over failure recovery process and can not ignore manual efforts to get the root cause analysis done; hence not being considered for improvement, currently
	Source Team Dependency	10	N	We don’t have any control over the source system and source team; hence not being considered for improvement, currently
	Ticket logged to Support Team. They Work as per SLA	10	Y

27 of 39

IMPROVE

28 of 39

Feasibility and Effectiveness Matrix

S.No	RootCause	Solution	Feasibility(F)	Effectiveness(E)	F*E
1	Current process consumes manual efforts in converting source feed into standard format	1. Build an UI tool which enables users to validate mandatory information in the Purchase Order raw Source feeds	2	4	8
1		2. Build a single common loading window for all the raw business source feeds which converts into standard format.	4	5	20

2	Manual Process consume more time in processing huge volume of Purchase Orders at one go.	1. Eliminate manual efforts by automating the validations	4	3	12
		2. Setup rejections check in prod prior to load to avoid loading in QA environment and build a auto correction setup for users to correct data	4	5	20
		3. Improve the performance of production loading window	1	4	4

3	Temp space Full/ Lack of temp space	1. Minimize the other business loads and utilized spaces as this is a shared environment	2	3	6
3	Temp space Full/ Lack of temp space	2. Set up a dedicated Database and server environment with enough space	1	3	3

4	Ticket logged to Support Team. They Work as per SLA	1. Dependency on support team should be eliminated	3	3	9
		2. Users should be able to trigger the load whenever they wanted with a UI tool	2	3	6
		3. Technical and Functional Training should be given to users and monitoring team	4	2	8

29 of 39

Action Planning

S. No	Root Cause	Solution	Action Plan	Responsibility
1.	Current process consumes manual efforts in converting source feed into standard format	Build an UI tool which enables users to validate mandatory information in the Purchase Order raw Source feeds	Set up an UI with which functional users can validate the mandatory details in the source PO feeds send by multiple business Once all PO information is available ,Users kick off automation ETL load which formats the raw file into standard format. Rejections in Purchase Orders are send to Functional users via mail. Once all rejections are fixed by users and amounts in PO matches ,users will kickoff ODS ETL load for GAP rules validation. After actual ETL Load completes, reports would be sent to Business for corrections in customer information if any. Users would be able to correct the customer related information by their own using UI tool. After successful load ,Purchase Orders are processed and sent to Finance team for calculation of incentives for GE sales executives	ODS DWH Dev Team
2.	Manual Process consume more time in processing huge volume of Purchase Orders at one go.	Setup rejections check in prod prior to load to avoid loading in QA environment and build a auto correction setup for users to correct data		ODS DWH Dev Team

Action Plan Matrix

Note: Currently, considering the F*E > 15 for action plan and implementation as these are having the maximum effect and controllable and achievable inside our project team’s scope.

30 of 39

Earlier vs. Improved Process

Improved Process Map

Using UI tool, functional users validate PO information

Purchase Orders loaded via ETL into CDW DWH

Reports are automatically sent to Users via mail

Calculation of Incentives for GE sales executives

GE Payroll team

Earlier Process Map

Raw Purchase Orders Manually processed

Formatted PO’s are loading in QA for validation

With an Approval, logged ticket to load in Prod

Purchase Orders loaded via ETL into CDW DWH

Reports are manually extracted and delivered to users

ETL load

Cognos Report Studio

Manual Process

Calculation of Incentives for GE sales executives

GE Payroll team

Cognos Report Studio

Rejections Fixed if any

31 of 39

FMEA (Failure mode and Effect analysis)

Process Steps	Potential Failure Mode	Potential Effects of Failure	Sev.	Potential causes of Failure	Occ.	Current Design Controls	Det.	RPN	Recommended actions	Responsibility	Status	Sev.	Occ.	Det.	RPN
Purchase Order Raw source feeds are validated using UI tool	Failure while uploading feeds into UNIX server	Impact in starting etl load if complete feed not transferred to server	7	Due to any network/environmental issue or any data issue	5	UI will upload the source feed to Server which triggers ETL load	4	140	Automated mail should trigger to functional users who uploads the source feed along with count of records	CDW DWH Dev Team	Implemented	4	1	8	32
Purchase Order Raw source feeds are validated using UI tool	Failure in timely recovery process	Impact in next incremental load schedule	7	Due to multiple team dependencies	3	Recovery window should be executed by Support team in case of any failures in servers/etl loads	7	147	Functional users get failure notification and they contact Support team for fixing the Production server issues	CDW DWH Support Team	Implemented	5	3	6	90
Data Loading in to Data Warehouse triggers one after other as per dependency Order	Failure during any of dependency load	Impact in next load	7	Due to any rejections, trigger file not available etc.	5	There will be a trigger file at the end of each load which kicks off next load	4	140	Automated mail should trigger to the respective support team members for resolution after every load to know the status	CDW DWH Dev Team	Implemented	4	1	8	32
Handle rejections and corrections	Any data issues	Data load will be impacted and users will be affected	7	Rejections/corrections in data	3	Extract would be sent to functional users for validation	7	147	If any corrections required, users should be able to correct in the same extract and trigger the load again.	CDW DWH Support Team	Implemented	3	2	5	30

32 of 39

Action Planning(Modified)

S. No	Root Cause	Solution	Action Plan	Responsibility	Target Date
1.	Current process consumes manual efforts in converting source feed into standard format	Build an UI tool which enables users to validate mandatory information in the Purchase Order raw Source feeds	Set up an UI with which functional users can validate the mandatory details in the source PO feeds send by multiple business Once all PO information is available ,Users kick off automation ETL load which formats the raw file into standard format. Rejections in Purchase Orders are send to Functional users via mail. Once all rejections are fixed by users and amounts in PO matches ,users will kickoff ODS ETL load for GAP rules validation. After actual ETL Load completes, reports would be sent to Business for corrections in customer information if any. Users would be able to correct the customer related information by their own using UI tool. After successful load ,Purchase Orders are processed and sent to Finance team for calculation of incentives for GE sales executives Auto mail notifications should be triggered at every stage with the status and schedule the triggers every 30 mins which enables users to correct data any point of time(Based on FMEA input)	ODS DWH Dev Team
2.	Manual Process consume more time in processing huge volume of Purchase Orders at one go.	Setup rejections check in prod prior to load to avoid loading in QA environment and build a auto correction setup for users to correct data		ODS DWH Dev Team

Action Plan Matrix (Modified)

Note: Currently, considering the F*E > 15 for action plan and implementation as these are having the maximum effect and controllable and achievable inside our project team’s scope.

33 of 39

Communication towards the Business Users

User Guidance

General Guidance to Business Users

# Guidance on UI tool is described on the instructions tab of the UI Macro which is user friendly.

# The CDW Data Warehouse is scheduled every 30mins which enables the users to rerun the loads in case of any rejections or corrections.

Note: Help Guide is provided to Users

CDW– Available Load Timings

Except ETL Dependant Load time:

# Dependant DWH DATA REFRESH would be done on Sunday which loads global customer data. So, functional users are not supposed to run during this time.

During ETL Daily Load /Outage time:

# CDW DWH would have planned and unplanned outages due to network/DB issues and users get notification not to run any loads during outage time.

Load Status mail notification enables users to have transparency on the loads running at backend ETL

Green –No loads in place – During scheduled loads of different dependant schema refresh.

Orange – During Outage time. Users not supposed to run to avoid load failures

34 of 39

Implementation Planning

Implementation Plan/Pilot Design

ERP source feed

Download source file to system

User runs required home appliance tool (Macro button) to convert source file to Unix compatible file

ETL /UNIX Scripts

AUTOMATION WORKFLOW

Converts to text tab delimited

Brings all different files

to common format

ODM

If no rejections,

then load in Prod

(ETL loads)

Extract sent to business for validation

CORRECTION LOAD

ETL Load to

pull loaded

Orders information

If any Orders need to be reprocessed

Users/Support team notified for rejections by emails

Orders are corrected in source files

Sign off

35 of 39

New Value Stream Map

Control

Step 10

Customer

I

PO’s are placed

in server and

Process is kicked

automatically

W

Purchase Orders

Are processed

if is valid, else

Users get alerts

With rejections

Reports are

Automatically

Sent to

Customers

W

15 Min

360 Min

5 Min

0 Min

PROCESSING

TIME

WAIT

TIME

Non value added activity

Value added activity

Value enabling activity

Team Member

Support Member

Functional User

Datawarehouse Load

If load

failures

Alerts via mails

36 of 39

Value Stream Before and After

Control

Step 10

Cost
Defects
Lead time
Waste

Cycle Time

Before

After

Identify

Raw feed

Screen

Door

Process

Manual

Processing

of PO

User

Verifies

POs

Provides

Approval

Ticket

Logged

For Prod

Ticket

Assigned

PO’s

Requested

Cycle Time

PO’s are placed

in server and

Process is kicked

automatically

Purchase Orders

Are processed if is valid, else Users get alerts

With rejections

Work Time (value add)

Wait Time (non value add)

Relentless Focus on Reducing non value adding activities

Min. 15 360 5

= Total 1460 Min.

= Total 380 Min.

QA

Testing

Reports are

Automatically

Sent to Customers

Same work completed in less time !!

Productivity = 1080 Min. Productive Time SAVED
Customer satisfaction = User gets report automatically.
Profit = Decrease in cost because of less developer intervention
Accuracy=Less intervention reduces human errors

Min. 20 180 270 35 345 45 10 20 25 495 15

POs

Processed

Reports

Delivered

Manually

37 of 39

Monitoring Techniques/Alert Mechanism

Can be monitored all the details of loads in Informatica Power Center Workflow Monitor tool.

Alerts at each stage to the Business, GAMS team and selected Project team members:

For Automation Load (conversion of source feeds to ODM format)–

Alert after ETL Load Trigger
Alert after First level rejections (With attachment of rejected records)
Alert after Second level rejections (With attachment of rejected records)
Alert after ETL Load Success

For Orders Load (loading of Orders information to DWH tables)–

Alert after ETL Load Trigger
Alert after session failure/sucess
Alert after ETL Load Success

38 of 39

Improved Process Acceptance

Customer Satisfied

39 of 39

Thank You!!!