Accelerating Outcomes in Big Data, IIoT/IoT, and AI/ML Talking AI with Hashmap July 2018
2 HASHMAP : WHAT WE DO We accelerate innovative business outcomes in BIG DATA, IIOT/IOT, and AI/ML with a combination of PEOPLE, PROCESS, AND TECHNOLOGY enabling END-TO-END SOLUTION DESIGN, DEVELOPMENT, & DEPLOYMENT
3 HASHMAP : ABOUT US Hashmap Corporate Snapshot HASHMAP QUICK STATS OUR SERVICES FOCUS AREAS Consulting Services Accelerator Services Managed Services – Founded in 2012 to provide Big Data Consulting – HQ in USA (Roswell, GA) with offices in Houston, - Big Data, IIoT/IoT, AI/ML - Ready-to-run, engineered, - and “As a Service” Canada, and India template-based approach Models - Domain Advisement, – Sole focus has always been Big Data & IIoT/IoT - Tempus IIoT/IoT - Architecture, Admin, Strategy, Architecture, –Business model is 100% consulting services and Data Engineering, Solution Optimization, - Cloud – Edge – ML software engineering services (we do NOT resell Consulting, and Automation, - Industrial, OPC, WITSML hardware, software, or subscriptions) Application Development Improvement –Subcontract our services - you make $$$ ATLANTA HOUSTON CANADA INDIA hashmapinc.com 1000 Holcomb Woods Parkway 24275 Katy Freeway 4145 North Service Road Midas Tower, Plot # 44, Building 100, Suite 118 Suite 400 2nd floor RGIP Phase 1, Roswell, GA 30076 Katy, TX 77494 Burlington, Ontario L7L6A3 Hinjewadi, Pune, Maharashtra, India
4 HASHMAP : ABOUT US 170 CONSULTING PROJECTS 125 COMBINED CUSTOMERS 1 24 INDUSTRIES Oil & Gas Financial Services ROW Technology Manufacturing Insurance Retail Pharma Healthcare Communications Power & Utilities Digital Media
5 PRESENTED BY DATABRICKS AT SPARK - AI SUMMIT 2018 Only a small % of enterprises are successful with AI & ML
6 PRESENTED BY DATABRICKS AT SPARK - AI SUMMIT 2018 Everyone Else Faces Major Challenges That Slow Down AI/ML Projects 1 DATA IS NOT READY FOR ANALYTICS 2 A ZOO OF AI/ML FRAMEWORKS 3 IT IS HARD TO PRODUCTIONIZE AI/ML
7 HASHMAP : CONSULTING IIoT/IoT and NoSQL Cloud and Data Data Science, Streaming & SQL Platform Engineering & Analytics, AI, & Analytics Services Data Pipelines ML Data Integration – Tempus IIoT/IoT Evaluation, comparison, Container Orchestration Rapid Analytics – – – – – Data Quality and Framework use case-based and Management Application Creation Curation Industrial control implementation DevOps Self Service ML – – – Spark Data Pipelines – systems integration Optimization, Tuning, Cloud Enablement and In Memory Analytics Grid – – – Data Warehousing, EDW – Time Series Dataset Performance Testing, Simplified Deployment Discovery & Exploration – – Enablement, Lake Shore Optimization and Benchmarking Data Virtualization Data insights – – Marts Edge Intelligence and Automated data Security Data Preparation – – – – Single View and – Edge Computing ingestion into Big Data Governance Predictive Analytics – – Customer 360 Deep expertise in Kafka platforms for SQL Operations – – Offloads and Data – and Spark Streaming datasets Performance – Archiving Data flow and data Low Latency Operational – – Enterprise Application – operations pipelines Data Stores Environments Batch and Real Time – Ingestion
8 HASHMAP : PROJECTS CHALLENGE APPROACH OUTCOME POC to determine marketing Automated social feed ad Provided quick visibility into content effectiveness using content for data acquisition, marketing content Natural Language data cleansing and curation, effectiveness through Processing ”NLP” in order to natural language processing, engagement prediction, post better focus ongoing visualization and data export clustering, and key word marketing spend and budget to provide content analysis, identification associated engagement analysis, and with engagement success spend analysis
9 HASHMAP : PROJECTS CHALLENGE APPROACH OUTCOME Integration of on-premise, Integrated real time data A production ready, secure real time streaming data collection and a high speed solution allowing the data sources including an message broker with the science team to quickly enterprise historian with on-premise data store operationalize and dynamic, predictive ML providing a dynamic productionize their ML models to provide insights streaming solution for models within hours to field operations automated ML pipelines
1 HASHMAP : OUTCOME & USE CASE ENABLEMENT 0 HASHMAP OUTCOME & USE CASE ENABLEMENT WORKSHOP UNDERSTANDING – COLLABORATION - STRATEGY – ROADMAP - ARCHITECTURE Attendees Onsite Agenda Key Dimensions Reviewed in the Business Context Segment Program Key Stakeholders (Business and IT) ⏤ Business Context – Detailed use case review ⏤ Organizational Effect and success criteria ⏤ Architecture Review ⏤ Business Impact Duration ⏤ Data Center / Cloud Review ⏤ Priority and Urgency ⏤ Ongoing Operations ⏤ Expected Time to Deliver 1-2 days onsite, 1-2 days offsite ⏤ Program/Project Time Horizons and Enabling ⏤ Risks ⏤ Innovation or Cost Savings Focus and Roadmap ⏤ Confirmation / Final Risk Identification / Impact Objectives ⏤ User Population and Consumption Deliverables Discussion / Wrap up DESIGN, REFINE, DEVELOP, DELIVER with an Patterns outcome-based focus ⏤ Capability Assessment and Review Deliverables Facilitate discussion and documentation of… ⏤ Datasets ⏤ Business Value Matrix and Detail for ⏤ Business Value ⏤ Other Dependencies Specific Use Cases ⏤ Solution Architecture ⏤ Technologies/Tools in use/planned/vision ⏤ Conceptual Architecture (Edge/Facility and ⏤ Ongoing Operations ⏤ ML/AI requirements Data Center/Cloud) and program/project ⏤ Timeline/Horizons ⏤ Analytics and visualization requirements horizon mapping Provide deliverables that… ⏤ Recommendations and Go-Forward Plan ⏤ Demonstrate business alignment ⏤ Outline initiatives, priorities, and time horizons ⏤ Highlight expected value
11 HASHMAP : OUTCOME & USE CASE ENABLEMENT
12 HASHMAP : LABS Review and Field Testing Edge Blockchain Industrial Computer Container Self Service Intelligence Enablement Deep Learning Vision Orchestration ML
13 HASHMAP : LABS CHALLENGE APPROACH In hazardous areas such as an Oilfield Using a convolutional neural network (image Drilling Rig, a Manufacturing Plant, or Heavy classification algorithm) applied to the whole Equipment facility, it is important to image at a single point in time, several frames understand the exposure of personnel to the can be analyzed per second. By splitting a video riskiest areas. Gathering this data is crucial image down into representative frames in real in understanding the exposure for time at the edge, we are able to transmit only the improvements in operational processes and events, rather than an entire image or video risk mitigation. segment. OUTCOME SOLUTION By applying real-time object detection, risk Utilizing a small edge device such as an assessments and mitigation can be Nvidia Jetson SoC (system on a chip) or a a undertaken before the occurrence of a ruggedized device, running a pre-trained significant safety incident. By including object-detection algorithm allows the safety related KPIs in a performance tracking identification of objects of interest and the program, organizations can have a holistic confidence of the identification. Based on a performance target in terms of both configurable threshold, events can be efficiency and safety. generated and transmitted over MQTT/CoAP or any appropriate IoT protocol.
Next Steps 1 2 3 Contact Hashmap Check out our Subscribe to our directly to discuss Engineering & Technology IoT on Tap opportunities to partner Blog Podcast Kelly Kohlleffel https://medium.com/hashmapinc https://www.hashmapinc.com/podcasts VP Sales & Marketing (713) 628-6030 kelly@hashmapinc.com linkedin.com/in/kellykohlleffel
Recommend
More recommend
