BrainDocs Solution Update Feb 2013
Download as PPTX, PDF2 likes2,088 views
The document describes a new product called ai-BrainDocs that helps legal, financial and compliance professionals build personal intelligent agents to identify relevant concepts within documents. The agents can save professionals significant time and money by reducing the number of irrelevant documents that need to be manually reviewed. The product is aimed at the multi-billion dollar eDiscovery and regulatory compliance markets.
BrainDocs Solution Update Feb 2013
- 1. The Solution for Experts Looking for a Competitive Edge February 2013 ai-one™ Intelligence delivered © ai-one inc. 2013 ai-one
- 2. Meet Your New Assistant(s) You train them, multiply them, share them. No overtime, no benefits, no complaints. © ai-one inc. 2013 ai-one
- 3. Quick Facts • ai-BrainDocs helps you build personal intelligent agents for finding concepts within documents in any language. • Customers are legal, financial and compliance professionals • Markets are multi-billion dollar eDiscovery and eGRCM (Governance, Risk & Compliance) • First customer shipped • Early Adopter Version Available Now © ai-one inc. 2013 ai-one
- 4. Big Idea Professionals armed with a personal intelligent agents they train to identify relevant concepts can save companies, legal firms and government agencies massive amounts of time and money. “digital data growth is explosive and digital data is the stuff of business and business disputes” - Gartner Magic Quadrant for eDiscovery May 2012 © ai-one inc. 2013 ai-one
- 5. What we do different Our solution is the ONLY one built with an ai-one “brain” (uses ai-Fingerprint technology) that addresses weaknesses of existing language tools, is language agnostic, works at the paragraph (concept) level and derives relevance from the context of use within the document. “Electronically stored information contains human language, which challenges computer search tools. These challenges lie in the ambiguity inherent in human language and tendency of people within networks to invent their own words or communicate in code.” - Best Practices Commentary on the Uses of Search and Information Retrieval Methods in eDiscovery, Sedona Conference © ai-one inc. 2013 ai-one
- 6. Why you need a BrainDocs Agent • We are an extension of YOU, the expert, not a black box replacement • We improve efficiency of manual but routine processes not addressed by other solutions • Built for lawyers, researchers, and analysts… not geeks • Execute your first project immediately on startup • Our technology engine is natively faster & more accurate © ai-one inc. 2013 ai-one
- 7. Customer-Problem-Solution Customer Problem Solution Expert legal, financial, Documents must be research, or read by experts and compliance they don’t have professional in solutions they can enterprise or professional services initiate, train and launch quickly and easily. Experts burn out reading thousands Personal intelligent of irrelevant agents can read documents and quality documents to flag suffers those needing review, eliminating wasted time © ai-one inc. 2013 ai-one
- 8. Everyone has an eDiscovery Problem • On average, employees generate 1 gigabyte of data per year. • If the allegations of a lawsuit involve 20 employees over a 10 year time period, you will need to collect and review for production to the adverse party 200 gigabytes of data…reduced to 150 gigabytes. • If it is assumed that each gigabyte contains 50,000 pages, there will be 7,500,000 pages for attorney review. • The claimed average review rate by law firms is 200 pages per hour; which breaks down to 37,500 hours of attorney time for the review. • If the market value for contract lawyers is $75 per hour the review will cost $2,812,500. source: A Kershaw Attorneys & Consultants © ai-one inc. 2013 ai-one
- 9. Cost Savings are Everyday • You don’t need a lawsuit to save money with BrainDocs, use it everyday to reduce your workload • Our testing shows we can reduce documents requiring expert review by as much as 50% • Value prop – for every $100,000 expert that spends at least 50% of their time reviewing documents, that’s $25,000 wasted on irrelevant documents… so ROI on BrainDocs is less than 60 days “The human review phase of eDiscovery is estimated to account for up to 80% of the total cost” - according to IDC 2010 © ai-one inc. 2013 ai-one
- 10. Benefits for User & Enterprise • Productivity- review more documents faster • Timeliness- faster project turnaround • Tighter compliance- risk mitigation • Relevant document accuracy • Higher job satisfaction • Cost effective on small projects • Perfect for eDiscovery service firms, enterprises and research organizations © ai-one inc. 2013 ai-one
- 11. Document Types | Processes • Engagement Letters • High Volume • Sales/Marketing materials • Operations Documents • Employment Agreements • Multi-Language • Non-disclosure Agreements • Compliance • Option Agreements • Review & Encoding • Leases • Manuals • SEC Filings • Surveys • Email and messaging • Free text in forms • Social media © ai-one inc. 2013 ai-one
- 12. Product Overview personal the analytics conceptual intelligent fingerprints agents we documents b ai-BrainDocs paragraph level storage concept discovery databases Intelligence discovered email content library • compliance • eDiscovery the brain ai-one NathanApp © ai-one inc. 2013 ai-one
- 13. Product Features 1. Documents to be analyzed are batched and imported into ai- BrainDocs case libraries (similar process to indexing), only once. 2. Agent(s) is created by user loading example paragraphs for concept “fingerprint” 3. User directs Agent(s) to analyze a library to rank by concept similarity score 4. User evaluates performance of Agent and continues teaching/testing or saves for production 5. Workflow queue is created and tagged documents are processed 6. User (Admin) customizable output with Excel or BI tools 7. Fully customizable UI/UX and database for workflow integration © ai-one inc. 2013 ai-one
- 14. Product Architecture © ai-one inc. 2013 ai-one
- 15. BrainDocs Workflow © ai-one inc. 2013 ai-one
- 16. BrainDocs Interface Simple User Interface- the agents are trained and libraries scored for further analytics and presentation or export © ai-one inc. 2013 ai-one
- 17. Agent Creation Input Fields for creating concept Agents Input Fields for known “always include” and “never include” words © ai-one inc. 2013 ai-one
- 18. Results – Table View Export options Files ranked by highest Columns concept display score document rank paragraph and link to the paragraph with highest similarity score © ai-one inc. 2013 ai-one
- 19. Results – Infographic View © ai-one inc. 2013 ai-one
- 20. Teach & Test Agents Quickly • Charts show teaching an agent starting with one example (sparse) and improving as more (14) examples are added to the agent • 200 (20 page) sales contracts were used in this case • Scores in “rich” case shows known target docs (black bars) isolated at top of list and no false negatives below 75% • Dynamic confidence color bands show user the improved accuracy as concept definition is enriched © ai-one inc. 2013 ai-one
- 21. The Money is in the Red Keyword and NLP fails with positives (black bars) throughout- you must read every document BrainDocs Agent shows known target docs (black bars) isolated at top of list and no false negatives below 75% © ai-one inc. 2013 ai-one
- 22. Key Metrics • Users create their own agents in less than an hour, needs less than 20 examples for training • Agents search for concepts in emails at rate of 3.5 million per hour • Increases productivity by at least 50% • Next release will Fingerprint a library at the rate of 150GB per day • Server Edition is $4,950 per year © ai-one inc. 2013 ai-one
- 23. BrainDocs Server Edition Features: • Concurrent Users • Batch Processing of Content Library: 1 • Agent Creation: 1 • Concept Similarity Analysis: 5 • Initial Fingerprinting time for new documents: approx 100k per day (per Enron email test) • Max Number of Documents in Content Library: No limit • Max Number of Agents: No Limits • Document Types: Microsoft Word, Adobe PDF (readable), Plain Text Hardware Software Operating System Processor: 2 x Intel Xeon CPU @ Microsoft .NET Framework 4 Windows 7 64bit (Personal 2.8 GHz Java SE Runtime Environment Version Evaluation Edition only) 7u6 (or higher) Windows Server 2003 64bit Memory: 8 GB of RAM Apache Tomcat Version 7.0.29 (or Windows Server 2008 64bit higher) Storage: ~ 30 GB Web Browser: • OS: ~15 GB • Google Chrome v21 (or higher) • Application & Server: ~ 5 GB • Mozilla Firefox v15 (or higher) • Remaining: ~ 10 GB to store • Internet Explorer v9 (or higher) content library (or higher if necessary) © ai-one inc. 2013 ai-one
- 24. If you’re ready to save time with ai-BrainDocs, let’s talk. Tom Marsh, COO ai-one inc. Follow us on Twitter @ai_BrainDocs 5711 La Jolla Blvd. La Jolla, CA 92037 Website www.ai-braindocs.com Ph: +18585310674 tm@ai-one.com © ai-one inc. 2013 ai-one