CloudFabrix - IT Press Tour #48 Jan. 2023

Transcript

1. 1 CloudFabrix Proprietary and Confidential. © 2015 - 2022 CloudFabrix Data-centric AIOps Platform Powered by Robotic Data Automation Platform Shailesh Manjrekar, Chief Marketing Officer
2. 2 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Key points we want to highlight! FY’22 onwards - Growth Phase • Market • Convergence around a single source of truth - New Business Models, Insights with Data • AIOps Operating Model • FinOps • Product • Robotic Data Automation Fabric • Low Code / No Code platform • Data-centric AIOps Platform • Observability Pipelines • Composable Analytics • GTM • Customer Case Studies • OEM traction • Channel, Marketplaces, and Developer programs
3. 3 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Company Background
4. 4 CloudFabrix Proprietary and Confidential. © 2015 - 2022 About Core Team 3 Successful Exits Built Multiple Enterprise Grade Platforms (Strong Enterprise Software Expertise) 10K+ Enterprise Customers Served Multi Billion Revenue Impact
5. 5 CloudFabrix Proprietary and Confidential. © 2015 - 2022 CloudFabrix - AIOps Leader & Innovator ➔ Founded in 2015 ➔ 80+ Employees ➔ 4th startup from founding team ➔ Recognized by Analysts - Gartner, Forrester, GigaOM, EMA... ➔ HQ – Pleasanton, CA, USA, India ➔ Sales Offices: US, India, Singapore, UK, Sweden, Israel ➔ Gigaom Radar 2022 ◦ CloudFabrix Leader and Fast Mover ◦ CloudFabrix Innovation Leader ◦ CloudFabrix AIOps for Edge vendor GigaOm 2021 AIOps Radar EMA Research 2020 AIOPs Radar Key Customer Engagements Key Partner Engagements GigaOm 2022 AIOps Radar Forrester Now Tech AIOPS - Q2’2022
6. 6 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Market Landscape – Product / Market Fit
7. 7 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Multi-Cloud Challenges Vendor Lock-in Control Costs Acquisitions Vertical Expertise Rise of Edge IT & OT Convergence Multi-Cloud Drivers Tools Sprawl 64% Say Observability & AIOps Difficult Differing API’s Infra, Cloud Services Differing Data Formats & Challenges Multi-Cloud Challenges 73% Enterprises using 2 clouds or more 81% Planning by 2024 $692B Cloud Market Size by 2025 23% Growth Rate in 2021
8. 8 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Static to Dynamic Zero-Trust Datacenter – “Butterfly Effect” Infra: PlatformOps, ITOps Networking NOCOps Security: DevSecOps Application: DevOps, BusOps Self-Service Personas Modern Hybrid & Dynamic Datacenter Infrastructure as code Identity based security Service based Networking Kubernetes CI/CD pipelines HashiCorp Terraform HashiCorp Vault HashiCorp Consul HashiCorp Waypoint, Nomad + Private Cloud Edge Cloud + ELK/ECS Lambda AKS/ACS Azure Functions GKE Cloud Functions Cloud Map App Mesh Proprietary Google Istio AWS IAM Azure AD GCP IAM Cloud Formation Resource Manager Cloud Deployment Manager Distributed Data Integration and Automation are paramount! 76% Respondents said they are using more than one cloud
9. 9 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Siloed Operational Domains have gaps Convergence around single source of truth – Business/IT, IT/Security, Security/Networking Observability Domain Specific & Tools Sprawl Alert Noise Data Challenges Security Full Stack Context Redaction, Enrichment Full copy in S3/Replay to SIEM/UEBA/SaSE Automation Remediation NLP Insights & Dashboards ServiceOps & ChatOps CloudFabrix’s RDAF Observability pipelines unify these Operational Domains
10. 10 CloudFabrix Proprietary and Confidential. © 2015 - 2022 ITSM Dev Teams Costs S/W Apps Cloud Engineering Platform Operations Engineering Operations Application Platform Traditional Operating Model AI/ML/ ITSM/ Automation Cloud Platform Engineering Engineering Operations Application Platform AIOps Operating Model Self-serve personas - DevOps Teams AIOps operating model for Platform Engineering Teams Rigid, Disjoint Business and IT Self-Serve, Data-centric, AI/ML based, Slower Time to Market, Slow Innovation Scalable, Faster TTM, Innovative Composable Analytics https://www.cloudfabrix.com/resources/aiops-operating-model-whitepaper/ AIOps Operating model is aligned for Platform Engg. and Self -Service Personas, however..
11. 11 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Competitive Landscape – Current State of AIOps
12. 12 CloudFabrix Proprietary and Confidential. © 2015 - 2022 AIOPs is Broken Data Automation and Full stack service mapping are the “Achilles heel” of AIOps - Rule Based Correlation - Longer time to learn Event aggregation / Correlation - Topology – 3rd Party - Tagging, Cookbooks Enrichment - Full Stack service mapping - No Metrics & Logs - No Remediation - No cross launch of tools ServiceOps – Virtual War Room Data Integration, Ingestion - PS Overhead - Data Gaps - Limited Predictive ML - Closed ML Predictive Insights - No Composability of Dashboards, Services, Pipelines and Bots Composable Analytics
13. 13 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Deliver true AIOps value-> After cfx CloudFabrix Observability Pipelines unify IT & Security, automate every step Low Code Bots based Data Integration and Ingestion Data Integration, Ingestion Self Service Persona based dashboard, Low Code Bot based Services, Search Pipeline, Routing Composable Analytics Automated Service Discovery, Dependency and Impact Map Enrichment - Full Stack service mapping Multiple layers of correlation - AI/ML, Stack, Time Based Event & Log aggregation / Correlation Recommendation Engine MELT visualization, Synthesizer, Bi-Directional Integration ServiceOps – Virtual War Room Built in MLOps, NLP for Continuous ML Predictive Insights • CloudFabrix unifies Observability, AIOps, and FinOps with - • Transformative patent- pending Robotic Data Automation Fabric, • Across Edge, Hybrid, and Multi-cloud Datacenters
14. 14 CloudFabrix Proprietary and Confidential. © 2015 - 2022 CloudFabrix Data-centric AIOps Platform “Robotic Data Automation Fabric is a Game Changer and Key Differentiator” – Forrestor, Gartner & GigaOm Analysts Powered By Robotic Data Automation Fabric - RDAF(™) 1000+ Bots Low-Code/ No-Code Composable Workflows Data Bots Distributed Data Fabric FinOps/ Asset Intelligence Log Intelligence AIOps Composable Analytics (Build your Own Service) Key Services
15. 15 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Business Overview - Vertical customers and Partners
16. 16 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Our Unique AIOps Story Major Customer Wins with expanding ARR Big Tech Partners betting on CFX AIOps Platform Growing Bots (1000+) Community & Ecosystem Major Solution & Reseller Partnerships Signed; Joint Customer Deals in Progress 250% YoY ARR Growth Expanding Stakeholders & ACV Building Data/MLOps Automation Ecosystem
17. 17 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Telco : CloudFabrix chosen over five other competitors Infrastructure & Applications – Large Telco provider responsible for 30% of the World’s Internet traffic MSS - VPN Software Stack CICD Application Stack • Business Problem – Apps/IT Operations, Engg. planning, Launch New Business Services • Solution - Cross-Domain Correlation, Alert Noise Reduction & Full stack Service Mapping, Topology: based on LRID model (Left to Right impact declarative model) • Benefits – 97% Noise Reduction, 60% + MTTI, 50%+ MTTR, Consumption Insights for Cross-sell and Upsell
18. 18 CloudFabrix Proprietary and Confidential. © 2015 - 2022 • Business Problem – Bring in Business Agility and accurate Decision Making ECC team has embarked on Digital transformation project and was evaluating various AIOps solutions Solution – Discovery of application/IT stacks, ML based correlation, RCA and Incident Resolution • Benefits – FinTech : Transforming AIOps with RDAF pipelines Incidents Remediation , Learning Full -tSack Discovery Data Enrichment X-Domain Correlation Predictive Analytics Service Management Data Integration Data-centric AIOps CMDB Exchange Financial Reporting Common Infra Config/ Assets Alerts, via WebHook Metrics via API alerts,, metrics via Kafka alerts via SNMP alerts via API json/csv files Alerts,metrics Via File CloudFabrix AIOps Unified event visibility across App, Systems, Network assets and tools stack 97.05% Event Filtering, Correlation and Suppression 550K+ Alerts->15K Incidents Context Aware Triage dashboards for Exchange (SCOM), Digital Next Gen (Splunk) and In View (ITRS) app incidents Reduced MTTI with faster Impact analysis (2+ hours -> under 30-mins) Reduced MTTR (4-hrs to less than 1-hr) with recommendation for Issue remediation and automation integrations (RPA)
19. 19 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Healthcare: CloudFabrix at Kaiser – 250K+ Assets Powered by Robotic Data Automation Fabric Data Fabric AI / ML Engine Asset Data Ops Data Biz Data AIOps IT Planning IT and Business Operations Asset Intelligence Service Real time visibility into Full stack interdependencies IT Change Refresh Impact Analysis Actionable Insights to adopt consumption based IT models Asset Intelligence Alert Noise Reduction Root Cause correlation Analysis Automated Incident Diagnostics & Remediation + AIOps service RDAF • 10+ Data centers • 250,000+ IT assets • 15K Apps • 20+ member team • Business Problem – expanding footprint across multiple markets. Pursuing digital transformation and technology refresh initiatives to provide best-in-class healthcare services for its customers • Solution – Asset Discovery, Application Dependency Mapping, Change Management, Utilization Reports, Root Cause Analsis and Incident Resolution • Benefits – 100% Asset visibility, 40% Reduction in OPEX, Per App Impact assessment from 240 Hrs -> 15 mins
20. 20 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Design Wins -centric AIOps Platform – Partners to replicate solutions across $B markets
21. 21 CloudFabrix Proprietary and Confidential. © 2015 - 2022 IBM Partnership • Global Consulting partner agreement completed • Part of IBM AIOPs everywhere campaign • Available in IBM sandbox environment • Joint Reference Architecture, with Instana and Turbonomics • Proof of Technology completed • Global Sales Enablement done • Completed update - 110 GTM and SA’s Improve TCO over 50% Productivity over 50% MTTI/MTTR over 60% Self-service, No-code In-place Analytics IBM QRadar IBM Instana IBM Turbonomics IBM CloudPak for AIOps A P I Trigg er Outp uts CRM, social Unified Data Collector Metr ics Eve nts L o g s Tra ces Applicatio n Depende ncy Mapping Cfx modules IBM Products BYOT/ D Bring Your Own Tool/Data IBM Instana IBM Turbonomics IBM Cloud Pak for Data CVE TIP Observability Data Lake AWS S3/ Partners/ IBM Coud Object Storage (COS)
22. 22 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Key Partnerships In Progress • Cisco FSO (Full stack Observability) + CloudFabrix AIOPs • AppDynamics • ThousandEyes • CloudFabrix AIOps • Completed • Evaluation • POC • Joint Demo Environment • Exec Buy-In • In AppDynamics Marketplace • InProgress • Channel Partner Enablement (ex: Presidio) • HPE Greenlake • CloudFabrix Robotic Data Automation Fabric for enabling MLOPs in Greenlake • AIOps solution In Greenlake fort Day-1 & Day-2 Automation Services • Completed • Technology Partner Program • Ezmeral and GreenLake testing • DXC Evaluated Several AIOps vendors and Selected CloudFabrix as AIOps partner • Completed • Vendor Evaluation • POC • Technology Partnership Agreement • InProgress • Targeting several enterprise customers in Q1
23. 23 CloudFabrix Proprietary and Confidential. © 2015 - 2022 FY’22 Technology, Channel and Developer programs validate value add Observability Data Lake and Security Data Lake Value added Resellers, Marketplaces Developer Programs
24. 24 CloudFabrix Proprietary and Confidential. © 2015 - 2022 CloudFabrix’s Data- centric AIOps Platform – Patent pending RDAF & services
25. 25 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Data-centric AIOps => Bringing it all together! Unified Events/Correlation ITSM Actionable Incidents Predictive Insights Predictive Alerting Incident Incident Analysis & Resolution Data Prep & Integration • Webhooks • APIs • Kafka • Streams • SNMP • Syslog • Splunk • Open Tele .. PIPELINES Full-Stack Service Map Composable Analytics Ops Admin Execs
26. 26 CloudFabrix Proprietary and Confidential. © 2015 - 2022 6 Differentiating pillars • RDAF Platform – Data Integration, Data Preparation, Eliminate Data Silos • Bridge Skills Gap – Low Code / No Code • Train and Retrain Models with Continuous ML – MLOps + Data Ops • AI Pipeline Synthesis and Recommendation Engine • Composable Dashboards, Services, Search and Pipelines • Full Stack Service Mapping and Enrichment
27. 27 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Full-Stack Service Map - Layers Example: Cloud-Native Hybrid Application Stack Business Apps & Components, App Host Inventory, Apps Topology RDA Bots & Pipelines Pods, Nodes, Service configuration RDA Bots & Pipelines Physical & Virtual Compute RDA Bots & Pipelines Tests, Agents, Devices RDA Bots & Pipelines
28. 28 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Demo Example Stack, Scenario & Personas ➔ Stack: Online Banking Application ➔ Scenario: DB Performance Degradation ➔ Impact: Application Page Load Times L1 Analysts Fewer incidents that are correlated and actionable L2/L3 Faster Root cause analysis & automate remediation Business Users Identify impacted elements for service degradation Application Application Infrastructure Shared Services Host/OS Infrastructure Monitoring Stack
29. 29 CloudFabrix Proprietary and Confidential. © 2015 - 2022 At-Scale Predictive Analytics
30. 30 CloudFabrix Proprietary and Confidential. © 2015 - 2022 3.Large Scale Predictive Insights: Incident Avoidance with Predictive Insights Prediction Cards Prediction Run on KPI: Interface Packet Discards Prediction Predicted Peak Anomalies Key Challenges: • Contextualize with Topology information – every deployment differs • Implement different approaches – Un-Supervised, Stack, Attributes Forecasting Accuracy >80% 90% Prediction Accuracy Training time <50% Key Benefits:
31. 31 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Using Cohorts 1000 Nodes 50 Metrics for Each Node 10 Cohorts for each Metric. 500 Models - Actionable - Scalable Sample Environment Cohort Creation & Data Reduction Rule / Attribute Based Topology Based Clustering Based Cohort 1 Cohort 2 Cohort 10 * Cohort rules can be tweaked to get right number of groups that are actionable
32. 32 CloudFabrix Proprietary and Confidential. © 2015 - 2022 FinOps – Asset Intelligence Analytics
33. 33 CloudFabrix Proprietary and Confidential. © 2015 - 2022 3. FinOps - Asset Intelligence and Analytics Real-time Visibility into IT Assets, Dependencies and their Lifecycle Data “ … saved us time & money… and greatly simplified the complexities of tracking Hardware/Software assets and their dependencies, ... we now have better control of managing lifecycle and upgrade initiatives … ” - Sr. Director, Cloud & IT Ops
34. 34 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Observability pipelines – Decoupling Producers and Consumers
35. 35 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Before : Existing solutions turning into Data Swamps
36. 36 CloudFabrix Proprietary and Confidential. © 2015 - 2022 After : cfxCloud Data Observability pipeline Powered by Robotic Data Automation Fabric
37. 37 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Observability Pipelines - Log Intelligence Service Benefits • Bring your own log tool (BYOL) • Ingest data in pull/push/batch modes • Up to 40-80% log volume reduction using correlation techniques • Replay using UTC timestamps, IP addresses, and certain patterns, to your choice of stream • Aggregate logs, normalize, transform, enrich • Route to multiple locations - data lakes, log stores, analytics platforms, composable dashboards • Enrich logs using Geo-IP or DNS looksups from InfoBlox, CVE MITTRE and TIP feeds • PII Mask sensitive information • In-Place Search, Collect and store only actionable data as a full- fidelity copy in Observability Data Lake • Replay on security breaches and compliance needs • Convert logs into metrics and use a number of regression AI/ML models for anomaly detection Log Ingestion Log Reduction & Replay Log Routing Log Enrichment, PII Masking Edge IoT, In-place search Log Predictive Analytics
38. 38 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Splunk – Before and After; 1 YR and 3 YR TCO with Log Intelligence TCO Savings $413K - $2.26M Correlation, pattern matching, Suppressing with Alert State Manager, AI/ML models TCO Savings $1.3M - $7M Correlation, pattern matching, Suppressing with Alert State Manager, AI/ML models
39. 39 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Log Search - Before: Existing Search Paradigm Collect + Forward all data Edge Collect + Forward all data Datacenter Collect + Forward all data Hybrid/Multi-Cloud Index all data - expensive compute for indexes Store all data on expensive storage Search with… Search Data-In-Motion In log stores like Splunk, Elastic, etc. Proprietary Query Language • Expensive flash block storage for random access • Storage for indexes • Replicate for DR • Only work with particular LogStore • No Composable Query Data Silos result in delayed insights No Universal Search Exorbitant TCO and Complex
40. 40 CloudFabrix Proprietary and Confidential. © 2015 - 2022 In-Place Search + Forward Edge Log Composable In-place Search After: New Search Paradigm In-place search -> Collect -> Store” Edge, data in motion, in an observability datalake, Timeseries and Log Stores In-Place Search + Forward Edge In-Place Search + Forward Edge In-Place + Route + Search + Store Edge Datacenter Hybrid / Multi-Cloud • Jupyter Notebook • Pipeline Builder • CfxQL CLI Source Bots Transform Bots API Bots Sink Bots Uniform Search & Query Data In-Motion RDAF Observability Pipelines Metrics - Time Series Database Search + Store + Replay Full Fidelity Observability Data Lake Search + Store + Replay LogStore - Splunk, ElasticSearch, SumoLogic, etc. Log Store Observability Data visualization Composable Dashboards • Faster Time to Insights and actions • Reduce Complexity and Cost • Remove data silos • Ease of use with Bots • Any data type • Datasets, Dataframes, Dependency Mappings, Service tickets, PStreams
41. 41 CloudFabrix Proprietary and Confidential. © 2015 - 2022 cfxCloud Search Bots RDA Bots Automate Data Ingestion, Data Processing and Data Routing Example Update Bots Source Bots Data Source Bots Retrieve data from a Data Source either in Streaming or Batch Mode Sink Bots Data Sink Bots Send data to data destination such as Data lake or other Endpoints Transform Bots Data Update Bots transform or update the input data and produce new data Updated Data • Filter Data • Select Subset of Columns • Compute New Columns • Fix Input Data • Aggregate Input Data • Enrich Input Data Using a Dictionary • Convert Input Data Types • Validate Input Data Model • Random Sampling of Data • Perform NLP Sentiment Analysis • Perform Unsupervised Clustering • Classify input Data • Predict Time Series Data • Classify using GPT-3
42. 42 CloudFabrix Proprietary and Confidential. © 2015 - 2022 RDAF Platform
43. 43 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Robotic Data Automation Fabric (RDAF) Low-Code/No-Code Bot Marketplace (Bot Library): 1000+ Bots Expand to Edge with Data Fabric Deployment Freedom Updates Actions AI/ML & Analytics Data Prep/ Transform Data Integration Filter Shape Aggregate Convert Mask ...
44. 44 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Self-Service persona based Outcomes Observability Pipelines for Data In Motion
45. 45 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Patent pending Robotic Data Automation Fabric Platform – Modular, Extensible Low Code Platform Observability Pipelines for Data In Motion
46. 46 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Explainable AI (XAI) Improving customer experience with AIOps using GPU acceleration Transparency Security Reproducibility Integrity Explainability • Unsupervised Learning • Supervised Learning • Reinforcement Learning • Federated Learning • Transfer Learning • OpenAI • IBM Watson • Hugging Face • GPT-3 and Gopher
47. 47 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Metadata Description, Data Filtering, Aggregation, Mapping Metadata/Data Shaping Data Masking, Data Encryption, Data Decryption, Access control, Auditing Data Security Data Signing, Data Hashing, Data Checksum Data Integrity Data Deduplication, Implode, Explode, Transpose, Mapping, Binning, Pivot/UnPivot Data Transformation Sample Data Management Ops you can do with Bots Files, Bulk/Batch data, Streaming data, Pub/Sub Topics, Bookmarking Data Delivery Data Dictionaries, Data Enrichment Data Quality Enhancement Generic REST client, API integrations, Named bots REST API/Integrations Data Simulation, Synthetic Data for Test/Dev, What-If Analysis Data Generation Data conversion, Data Formatting, Data Templating Data Formatting Data Lineage, Data Tracing, Pipeline tracing, Data Version Control Data Governance
48. 48 CloudFabrix Proprietary and Confidential. © 2015 - 2022 Thank You! Free Signup for cfxCloud cloudfabrix.com/saas-login
49. 49 CloudFabrix Proprietary and Confidential. © 2015 - 2022 cfxCloud Roadmap – Ecosystem and Solution enablement CY’23 Continuous Innovation Edge AI and 5G/ Industry 4.0 use cases IBM and AWS AIOps Integration Cisco FSO Cloud Integration • • Splunk Cloud Integration AWS APN Sales Enablement and Motion
50. 50 CloudFabrix Proprietary and Confidential. © 2015 - 2022 AIOps at OTP Bank Webhook, API Web hook Web hook Incients (Manual) Webhook, API vROps APM Command Center Manual Review ➔ 1,500+ Branches, 2DC ➔ 10,000+ IT Assets ➔ 60,000 alerts/day ➔ 800+ Ops Team/24/7 ➔ Manual review/ticketing 800+ Ops/Service Desk Team Webhook, API Webh ook Webhook Webhook, API Auto Incidents vROps APM Command Center CloudFabrix Data-Centric AIOps Largest commercial bank of Hungary. One of the largest independent financial service providers in Central and Eastern Europe ➔ Data integrations: 100% ➔ Noise Reduction: 45% → 80%+ (over 3-months) ➔ MTTR: 2+ hours → Under 30-mins ➔ L0 Automation: 100% ➔ Opex Cost Savings: 40% (over 18-months) Key Outcomes