Rapid Analytics A visual, live approach to requirements gathering and business analytic development Mark Marinelli, VP of Product Management Brought to you by:
Agenda Why Do Traditional Analytics Projects Fail? What Is Rapid Analytics? Demonstration Rapid Analytics Opportunities In The Business Conclusion Q&A 2
Lavastorm Analytics at a Glance We give enterprises a new, agile way to analyze, optimize and control their data and processes. - 15+ years in operation - Major offices in USA, UK, Australia - Thousands of users at 80+ customers - 120+ employees - Billions of dollars in customer value created Lavastorm’s software makes business analysts heroes by giving them the power of a programmer to: • Rapidly unify disparate data • Easily construct complex analytics • Effectively deliver actionable insights and results CUSTOMERS (partial list) PARTNERS AWARDS
The Problem … 70-80% of Corporate BI Projects Fail 4
The Main Cause of Failure – An Understanding of the Real Requirements Sources of Failure Poor communication between IT and the business Failure to ask the right questions Failure to think about the real needs of the business Requirements Application 5
Traditional Approach and Technology for BI Thrives Where Requirements Are Slow to Change Traditional BI Providing Best applied where: Uses • Stability Reigns • Reports and Data sources & requirements dashboards with Relational change little answers to pre- DBs determined questions Cubes • Data is well known, well • Analysis of highly controlled and modeled and • Technologists answer most, cleansed data Requirements if not all, the questions Design • Very limited, pre- Implementation • IT owns all of the source determined drill paths data or very complex Verification query tools for Maintenance technologists • IT development and governance
But the BI/Analytic Needs of Businesses Have Changed Business Compete on Analytics Customer Expectations Speed, Accuracy, Visibility Speed, Specialization More Complexity Big Data More data, decision makers, change (ex: N-tier Supply Chains, Fractured Value Chains) Volume, Variety, Velocity 7
Rapid Analytics Is A New Approach Requiring Different Technologies A New Methodology Enabled By New Technologies Requires new technology that Emphasis on rapid prototypes , not enables agile methodology – traditional specifications choices include self-service BI, data discovery, mashup options Emphasis on direct business end user interactions rather than using Data-driven , not schema-driven, documentation or working through IT models and analytics liaisons Ability to handle complex, disparate, dissimilar data structures Emphasis on responsiveness and reactions , not fortune telling or time- Easier UI for end user self-service consuming over-engineering 8
Lavastorm Analytics Engine Overview Rapid Agile Extract, Transform, & Load • Merge any data (e.g. RDBMS, Excel, ERP / CRM, Analytics XML, binary, etc.) – no code required • Big Data ready Driven By a • Profile, inspect, cleanse, and transform sources Unique ETL Combination of Capabilities 5GL BI 5 th Generation Language Business Intelligence/Analytics • Process modeling • Self-service control for discovery/ad hoc analysis • Not traditional programming • No IT involvement • Rapid visual modeling using pre- built “nodes” • Data visualizations • Reusable, sharable components / models • Automated continuous monitoring/auditing 9
Demonstration 10
Problem Description Telecommunications operator Goal: Every customer is billed for their services – Analysis necessary to determine where unbilled customers exist, and optimally why their records are not accurate Challenges: – Diverse data - multiple acquisitions have produced a mix of multiple billing systems and a heterogeneous network – None of the relevant data sources are warehoused, but extracts are produced for other purposes – No visibility into the problem, so no business case to consolidate data sources into the EDW – Small team will research and correct the errant account records, so cannot waste effort on false positives
What Did We See In the Demo Incremental, discovery-based analysis – Real-time, collaborative requirements->implementation->validation process using intuitive visual tools Accurate answers in seconds – Business rules are tested with live data, with validations and improvements applied in real time – Rapid, low-cost application of hypotheses encourages the search for the exceptions and edge cases which often account for process infirmities Rapid data acquisition – Initial data sets obtained from raw sources without the need for intermediaries, structured schema, or warehousing Self-sufficient data federation and analysis – Ability to manipulate and prepare data sets within the tool fosters independence from traditional data brokers – Tools for testing data integrity avoid pollution of these satellite data sets 12
The Benefits of An Agile Approach – Examples from the Field New Data Sources 400 hours invested with no results using traditional means, but 3 days to a working application with Lavastorm - Communications Service Provider Greater Responsiveness “Sometimes I can get the information I need while I’m in the middle of an IM with somebody, or, if they’re in the meeting, I may have the information before the meeting is over…it almost makes people a little shocked that we can get the answer that quickly .” - Tom Tannehill, Lead Analyst, CenturyLink More Visibility Went from ignoring information because no business case to warehouse the data to uncovering tens of millions of dollars of lost revenue - Consumer packaged goods company 13
Opportunities for Rapid Analytics Are All Around In Departmental and Corporate Boundaries 3 rd Parties, Partners, Competitors Customers Agile Agile Research, Support, Agile Agile Agile Agile Manufacture Procure Market, Sell, Deploy Design Distribute Agile Hire, Manage Agile Finance
A Rapid Analytics Solution Can Sit Alongside & Complement Your Traditional BI Infrastructure Data Individual Data Process • Data discovery Controls • Continuous analytics • Rapid prototyping Data / BI Infrastructure • Data integration and management • Visualizations Data DWs Marts Ent. Apps Cubes Dashboards, Reports Data
What Does It Really Mean? Rapid Analytics is an entirely different experience for IT and information consumers All enabled by new technology that is complementary to traditional BI User experience – Analytic applications, not reports – All relevant data is joined and in one place – Real-time updates are possible – Changes are expected – Designed for business users, no IT degree required IT approach – Data foundation build through light mapping, not heavy modeling – Work with data “as is”, no cleansing required – Collaborate with the business throughout Opportunities for Rapid Analytics are all around and across the company – where speed matters
Thank You, Next Steps Get Lavastorm Analytics Engine Public Edition (FREE) http://www.lavastorm.com/softwaredownloadsandtrials Contact Me Follow Us Mark Marinelli Lavastorm_News Lavastorm Analytics Lavastorm Analytics Group +1 617-345-5422 Lavastorm Analytics mmarinelli@lavastorm.com www.lavastorm.com Brought to you by:
Recommend
More recommend
