CFO of Tomorrow Changing CFO capabilities in a world of big data

Transcription

1 Office of the Under Secretary of Defense (Comptroller) Office of the Deputy Chief Financial Officer CFO of Tomorrow Changing CFO capabilities in a world of big data Greg Little OUSD(C) Business Integration Office Graham Evans Booz Allen Hamilton

2 A centralized, standardized CFO function will provide real time insights 2

3 Hype vs. Reality? 3

4 Concept of data analytics is not new CFO act of 1990 empowered modern govt. CFOs to create enterprise-wide analysis Purpose of the modern CFO's office 1. Bring more effective general and financial management practices to the federal government 2. Provide for improvement of accounting systems, financial management practices, and internal controls 3. Provide for complete, reliable, and timely data to better finance, manage, and evaluate government programs 4

5 The Year 1990 #1 Song in 1990 #1 Movie in 1990 Technology in

6 What is different this time? Strong demand from leadership Audit Technology We are actually seeing value NFR tracking tool impacts all UoT first focus on 4 th Estate Cost management impacts all 6

7 UoT Existence and Completeness Flow 7

8 Data architecture: Centralized architecture will enable better governance and automation Current DOD system landscape Future state use of UoT Distributed data and management environment Some integration, but lacking visibility to full spectrum of DoD data Insights limited to single line of business or organization Siloed data Retention of current disaggregated system structure UoT collects data from multiple sources Creation of a "Rosetta Stone" to allow for enterprise-level views Enterprise view with UoT By Service Key systems Cross-Service Supplemental systems UoT Army GFEBS ISR MEPRS DMDC DataStar RFMIS infads Reporting ISR Cost SNaP-IT Contract Library Air Force CRIS DEAMS NAVAC Bud. Est. NAVAC. Brief DEP ARC Bud. Est. Data Warehouse MHS MDR DMDC Navy Marines Navy ERP FIS STARS-FL SABRS MHS MDR DMHRSi NAV-ITAS Contract Library FIDS DLA PBIS-IT SNaP-IT FPDS Aero. Evacs PCOLS Non ERP System ERP MEPRS CCR RFMIS GFEBS DMHRSi FPDS PCOLS DLA DoD data source system 8

9 Use Case 1: Reconcile Can we trace this? 9

10 Use Case 1: Reconcile When the following is true? All of these systems, plus 15,000+ rows of crosswalks support FBWT 10

11 Learning from Reconciliation We can learn about system issues and drive change through specific problem statements uot_sglprefix uot_signedamount uot_transeffdate uot_transid uot_transpostdate Posting date is almost three years after effective date fund application fund master uot_begfy uot_endfy uot_transpostdate Master data changed after posting creates reconciliation issue supplier uot_fednonfedind Uot_signedmount 021-W*** USA *** MAT CTR EUR***** N $ (60,273.50) Federal vendor rolling up to non federal payables Controls can always be improved, because people will always find a way... 11

12 Use Case 2: Performance Are we getting better? How do we know? 12/18/15 Appropriations Enacted Public Law No: A What types of obligations were executed in the final month of fiscal year 2017? 12

13 Learning from financial metrics A ENTITY obligated 24.29% of O&M budget in final month % of the are contract obligations for information technology and professional services Several of those obligations are with suppliers who have high dollar sole source thresholds Behaviors will change when we are able to trend with these details 13

14 Cost Management - Real Property: Key themes and opportunities Wide variation in electricity spend Estimate potential savings through establishing and managing to cost performance baselines Significant lease spend near existing DoD facilities Leased administrative space is within a 30-mile commute of dense DoD footprint Identification of regional clusters Potential for CODE analytics to inform regional consolidation Above: Electricity cost per square foot plotted for 400+ installations across all Services (each dot is different installation) Above: Administrative leases arrayed according to cost and amount of nearby owned/leased administrative facilities Top: DoD sites with administrative square footage. Bottom: Top 20 high density clusters of administrative facilities 14

15 Cost Management Real Property: CODE gives Installationlevel cost comparisons across all categories of spend Service-level spend and installations considered in-scope Specific installations Categories of spend (ex: construction, utilities, custodial, operations, etc.) 15

16 Cost Management Cost comparison of real property Cost comparison between AFB A, portfolio medians, and commercial reference estimates: Air Force Base s A real property costs in FY 16 were broadly comparable (on a per sq. ft. basis) to portfolio medians; however, cost appeared high in both Electricity and Sustainment 16

17 Cost Management Electricity Spend at Air Force Base A Electricity cost per square foot comparison vs. CORE benchmarks: Air Force Base A Cost Analysis: Total FY 16 Electricity Cost* Total Square Feet Cost per Square Foot ~ $7.5 Million ~ 3.44 Million Square Feet $2.18 / Square Foot Cross Service Regional Median: $1.12 / Square Foot Next step: Review and identify best practices being employed at installations outperforming commercial reference model 17

18 Cost Management - Information Technology: Key themes and opportunities Higher spend compared to Commercial References Delta with Commercial Reference model, driven primarily by: External services IT personnel Lack of consolidation in third party vendors High concentration with top vendors, with long tail of spend Inefficiencies in usage of external service providers IT spend/fte increases in MAJCOMS that have greater % of spend on external services 18

19 Insight to Action This all means nothing, unless Local Willingness to make reconciliation a habit Willingness to change processes Shared Service Willingness to adopt culture of data Willingness to take calculated risks System Level Willingness to make system changes Willingness to make agency performance a priority 19

20 CFO of Tomorrow: OUSD(C) Business Integration Office is becoming a Value Added Shared Service Provider Evolving the role of the DOD CFO to a Data Driven Performance Management Organization Auditor/ Accountant Budget Analyst Executive Portfolio Manager Insights AUD-IT AS A SOFTWARE & SOLUTIONS PROVIDER Pre-Canned Dashboards Self-Service Analytics Ad-Hoc Query We Develop: Applications Dashboards Interfaces We Provide: Training Onboarding User Forums Data Sources AUD-IT AS A DATA PROVIDER We Maintain: Automated Data Pipelines Analytic Queries Data Access Solutions Custom Data Views Accredited Infrastructure 20

21 CFO as a Value Added Business Partner FM Solutions Provider DESCRIPTION The delivery of insights derived by FM analysts and data scientists to assist customer support their mission Financial Statement Preparation Audit Workbooks Custom Dashboards SERVICES PROVIDED Performance Management Consulting Custom Application Development Software and Tool Provider A suite of web-based FM analytics tools fully integrated with the AUD-IT data platform removing the need for customers to purchase, install, and run applications on their machines Intelligent Data and Analytic Discovery Portal Self-Service Analytics Dashboard Templates Analytic Applications Virtual Desktop for Analytic Tools Data Provider A scalable single source of truth holding both curated and raw financial data allowing customers to focus more on the discovery of insights to support mission objectives Guided Data Ingest and Preparation Ad-Hoc Query and Share Data Catalog Data API Data Delivery Platform Provider Provides a common stable and secure platform for shared analytics. Alleviating customers from tasks surrounding security, performance, identity management, and compliance Identity and Security Measures FISCAM Compliance ATO NIPR/SIPR Environments Network Monitoring 21

22 Design Challenges Starting with the Audit Use Case MISSION & PROGRAM The lack of a universe of transactions is the #1 audit deal breaker for the 4th Estate Office of the Undersecretary of Defense Comptroller OSD(C) TECHNOLO GY Myriad Non-Standard interfaces, outdated tools Require scale, performance and low cost Auditor/ Accountant DATA & ANALYTICS Significant effort required to derive data insights Slow/no adoption of advanced analytics capabilities Data Sources 19 Accounting Systems 50+ Feeder Systems 4 Financial Statements SECURITY & AUTHORIZATI ON Data aggregation across disparate financial systems Compliance with DoD Security and FISCAM Controls Scale 35 Appropriations 24 Entities +Billions Records 22

23 Analytics Platform Architecture Users are presented with visuals from a variety of Business Intelligence tools and custom dashboards Power users can create visuals and explore data with native tools TOOLS AND CAPABILITIES Multiple tools allow users to explore, enrich and aggregate data as well as perform advanced calculations Analytics execute fast on large data sets with the use of high powered hardware Data is mapped to a common model for ease in data discovery, governance, and access Finding and accessing relevant data is simple COMMON DATA PLATFORM Data is stored securely in a variety of database structures to support Advanced Analytics at scale Data is secured through Role and Attribute Based Access which supports auditability and governance Data pipelines are automated against established business rules Data ingest is transparent and controlled Infrastructure is modular and scalable for maximum agility and performance Advanced encryption provides the basis for a robust security model 23

24 Design Challenges: Demystifying Big Data BELIEF If I combine all my data together I will have full insight into what my organization is trying to accomplish If I leverage industry technologies, I will be successful Once I have all my data in one place, I can focus on analytics There aren t enough data analytic experts to ask the right questions vs. REALITY Policies, restrictions, and security need to continue to be enforced and the original origin of the data and any changes to it need to be tracked throughout the life of the data asset Successful organizations leverage industry technologies, but they also adopt an agile culture that enables them to try new ideas and pivot quickly based on lessons learned Stored data needs to be cataloged to enable analysts. Information such as data set name, data formats, data tagging, release-ability, retention, and other metadata must be available. Data driven solutions need to create capabilities that democratize analytics to a wider audience The more data I have, the more questions I can answer, the more knowledgeable my organization will be It s about the right data, combined with the right tools, for the right problem not how much data is stored Through 2017, 60% of big data projects will fail to go beyond piloting and experimentation, and will be abandoned. Gartner, 24

25 Turning disparate data Into an Integrated Knowledge layer DATA SOURCES SOCIAL MEDIA NEWSFEEDS COLLECT Get and track data from the source PROCESS Give Data the Power of Greater Context AGGREGA TE From disparate data sources, one version of the truth EXPOSE Abstract away complexities through a single interface VISUALIZATIONS APPLICATIONS DATA SCIENCE ANALYTICS WEB CRAWLERS PROPRIETARY SOURCES Security, Governance, Provenance, Lineage Insights BUSINESS INTELLIGENCE REPORTING DATA CONSUMERS 25

26 Big Data technologies and design patterns offer a new paradigm for transitioning to a data centric organization Massively scalable ingest, storage, and processing using distributed, parallel processing and storage, so you throw away nothing Proven, mature open source software applications developed and used by some of the largest organizations in the world (Google, Amazon, Facebook, etc) On-demand expansion using built in Netops tools enable rapid, on-the-fly deployment and configuration management of bare metal hardware nodes Multiple analytics packages and methods supported clustering, data mining, in-memory analytics, pattern analysis, anomaly detection, network analysis, etc. High availability built into the architecture distributed storage and processing eliminates the need for complex SAN infrastructure and backup/recovery support 26

27 Big Data Architectures Allow for Sandboxes and Scale at Fractions of the Cost Data Governance Data Catalog Data Quality Data Provenance and Lineage Data Sharing Services Search Query Mutation Services Enterprise Application Integration Publish/Subscribe External Data Connections Business Process Management Data Security Cell Level Security Row Level Filtering Attribute Security Centralized Policy Enforcement Analytic Data Zone In-Memory Optimized Indexes Graph Analysis Data Warehouses Machine Learning Sandbox Data Zone Combination of trusted and raw data Data Sampling to support Development, Experimentation, and Prototypes Trusted Data Zone Master Data Cleaned, Processed, Organized Correlated Entities Policy Enforcement Version Controlled Data: Streaming Batch Structured Unstructured Raw Data Zone 1. Collect 2. Prepare 3. Organize 4. Profile 5. Enrich 6. Aggregate Unaltered Source Data LDAP / ADFS Establish data policies Cleanse and normalize data Extract metadata Store and catalog Data Archive Assess Data Quality Manage Version Control Track provenance / lineage Enhance metadata Create relationships Combine disparate data sources, create single version of truth 27

28 Enabling Self-Service Analytics through Data Discovery Data Provider Software and Tools User Starts With: What questions do I want ask of my data? IT enables answering the tough questions: 1. What data do I need to answer my business questions? 2. Where is the data that I need? 3. What does this data mean? 4. How do I access the data I need? 5. Can I trust the data available? Automated Data Profiling and Tagging Configurable, SLA based, Data Pipeline with conformance to Common Data Model Automatic data tagging, profiling, and lineage mapping on ingest Searchable Data Catalog Customized and standard metadata model for quick search and workflow automation Easy view Data Lineage Reporting from raw to curated data sets depending on use case Application Integration Robust API conforming to data aggregation security parameters Centrally managed authentication for custom app development or access Data Strategy/Data Governance Rigorous CM of data model and reporting on data drift End-to-End Data traceability for FISCAM reporting and audit controls Robust community for establishing best practices and MOAs Dashboard Pre-canned analytic dashboards supporting trusted analytical models for Financial Analytics Custom, use case specific app development using industry leading tools such as QLIK Workbench Suite of tools accessing a secure API, enabling, ad-hoc analytics Perform transformation, data wrangling, data science on raw and curated data Data Portal Preform Large File ingest with automated data checks, tagging, indexing and notifications Monitor data ingest and egress across the platform according to user profile and job ID 28

29 Questions 29