GraphSummit Singapore Master Deck - May 20, 2025
0 likes445 views
GraphSummit Singapore Master Deck - May 20, 2025
![Conditional Queries
provide more flexibility and expressive
power to handle complex querying
scenarios
Property Access control (PBAC)
data-driven rules to control READ and
TRAVERSE privileges on nodes
ownership, contextual access control
NEW CAPABILITIES
Cypher
Improvements
GRANT privilege property
ON GRAPH {database|*}
FOR ( var:label )
WHERE condition
TO role
GRANT privilege element
ON GRAPH {database|*}
FOR ()-[ var:type ]-()
WHERE condition
TO role
MATCH (n:Person)
CALL (n) {
WHEN n.age > 60 THEN {
SET n.ageGroup = 'Veteran'
}
WHEN n.age >= 35 AND n.age <=
59 THEN {
SET n.ageGroup = 'Senior'
}
ELSE {
SET n.ageGroup = 'Junior'
}
}
RETURN n.name AS name, n.ageGroup](https://image.slidesharecdn.com/forexternalsharinggraphsummitsingaporemasterdeck-may202025-250521084659-74c95970/85/GraphSummit-Singapore-Master-Deck-May-20-2025-96-320.jpg)
GraphSummit Singapore Master Deck - May 20, 2025
- 1. Welcome to
- 3. Agenda 09:40 Keynote: Graphs + AI: Your Enterprise Advantage Sudhir Hasbe, Chief Product Officer, Neo4j 10:10 Unlock the Power of GenAI Tim Sheedy, VP Research and Chief AI Advisor, Ecosystm 10:30 Morning Break 10:50 Harness Graph Power with AI: An Insights Discovery Use Case Lois Ji Ronghui, Data Scientist, AI Practice, GovTech (Skillsfuture Singapore Forward Deployed Team) 11:20 Neo4j: Product Vision and Roadmap Michael Hunger, VP -Product Innovation, Neo4j 12:00 Q&A Panel 12:30 Networking Lunch Reception
- 4. Agenda 14:00 Hands-On Workshops Workshop A: Create a Graph-Backed App from Scratch | Vista 2 & 3 Xavier Pilas, Sr. Solutions Architect, Neo4j Bryan Lee, Consulting Engineer, Neo4j Workshop B: Building Smarter GenAI Apps with Knowledge Graphs | Grand Ballroom Siddhant Agarwal, Tech Community Manager, Neo4j 15.15 Afternoon Break | Tea & Coffee 15.30 Hands-On Workshops Continue 16:30 Networking Reception
- 5. Graphs + AI: Your Enterprise Advantage Sudhir Hasbe | Chief Product Officer, Neo4j
- 10. The power of the graph model
- 11. The power of the graph model
- 14. Network & Security Product Customers Transactions Process Suppliers Graph is insightful Employees Network & Security What’s important? Process Suppliers What’s unusual? Product Transactions Customers What’s next?
- 15. Talent learning & development Career management Skills management Orgs (who is who) Job search Account/identity control Reputation scoring Threat detection Access control Zero trust Route planning & optimization Real-time shipment tracking Inventory planning Risk analysis Product recommendations New product introductions Product customizations Product inventory Product pricing Bottleneck identification Process improvement Process automation Process monitoring Recommendations Loyalty programs Churn prevention Customer offers Dynamic pricing Intelligent ads Anti-money laundering Circular payments Fraud detection Network & Security Suppliers Product Customers Finance Process Connect your organization into graph to drive transformation Employees
- 16. 16 Trusted by 84 of the 8 / 10 Top retailers 10 /10 Top automakers 10 / 10 Top US banks 9 / 10 Top aerospace & defence 9 / 10 Top telcos 10 / 10 Top technology & software 8 / 10 Top insurance companies 9 / 10 Top pharmaceuticals
- 17. NOW Connected data powers transformation It's essential for enterprise AI
- 18. Neo4j Inc. All rights reserved 2024 Generative AI races toward $1.3 trillion in revenue by 2032 Generative AI Revenue Generative AI as a % of Total Technology Spend However, 71% of organizations are stuck piloting GenAI projects 2024 IBM CEO Survey Bloomberg Intelligence
- 19. 85% of IT leaders cite data mgmt as their top AI challenge KPMG AI Quarterly Pulse Survey 2025 100 senior executives from $1B+ revenue companies It’s a data management issue! Data Organization Lack of Standardization Scale of Query Diversity AI needs data organized in ways it can effectively access AI requires high-quality harmonized data to be reliable AI demands rapidly accelerated question-to-query execution
- 20. Flexible Standardized Enables AI to grasp meaning and access data across sources Supports diverse query types and evolving data needs Maintains consistent understanding at scale What makes data AI-ready? Data exposes entities and relationship understanding Adapts to new patterns and structures Consistent, governed data foundation Contextual
- 21. Entities and their connections are built into the data model Evolve your model while queries keep running Consistent meaning across all your data sources Knowledge graphs = AI-ready data Relationships are treated as primary data points Schema grows and adapts to meet evolving business needs Expressive data model provides a common language Standardized Contextual Flexible
- 23. GraphRAG is RAG where the R path includes a knowledge graph. What is GraphRAG?
- 25. Vector Index built into the database Store any property, node, and relationship as vector in the database Ability to create embeddings by directly calling various embedding services like OpenAI, Azure OpenAI, VertexAI, and Bedrock Vector Support NEW CAPABILITIES
- 26. Integrated with GenAI ecosystem Support for all major frameworks like Langchain, LlamaIndex, Spring, Haystack, Semantic Kernel, etc. Integrated with all LLM platforms like AWS, GCP, Azure, and OpenAI GenAI Ecosystem NEW CAPABILITIES
- 27. GenAI Language Statistics Creativity KGs Knowledge Facts Context Demo: Knowledge Graphs Unlock GenAI
- 28. Benefits of 28
- 29. 1. Higher accuracy GraphRAG Performance RAG Performance Lettria Analysis1 81.67% 57.50% Writer Knowledge Graph2 (RobustQA Benchmark) 86.31% 32.74%–75.89% RAG vs. GraphRAG: Multi-hop Question Answering3 77% 66% GenUI Experiments (MultiHop-RAG Dataset)4 Successfully answered complex, multi-step queries Struggled integrating data from multiple sources GraphRAG delivers up to 3x higher accuracy than traditional RAG, with better multi-hop reasoning for context- rich AI applications. 1) https://writer.com/engineering/rag-benchmark/ 2) https://www.lettria.com/blogpost/vectorrag-vs-graphrag-a-convincing-comparison 3) https://arxiv.org/abs/2502.11371 4) https://www.genui.com/resources/graphrag-vs.-traditional-rag-solving-multi-hop-reasoning-in-llms
- 30. 2. Easier development Natural language: Apples and oranges are both fruits…
- 31. 3. Improved explainability X Customer Service Doctor Social Security Number Social Security Number Patient Bob Phone Number Health Diagnosis
- 33. What does this look like in 33
- 34. Internal Documentation Wikis Enterprise Systems Klarna transforms knowledge access with GraphRAG HR Systems
- 35. Internal Documentation Wikis Enterprise Systems Klarna transforms knowledge access with GraphRAG HR Systems Daily queries processed 250K Employee questions answered in first year 2,000 85% Employee adoption
- 36. Gaming leader transforms analytics with GraphRAG Game Data Sales Figures Customer Feedback Gaming Leader Marketing Data Enterprise Data
- 37. Gaming leader transforms analytics with GraphRAG Game Data Sales Figures Customer Feedback Gaming Leader Marketing Data Enterprise Data Reduction in time spent on data requests 10x Time-to-insight reduction 92%
- 39. Agentic RAG is an evolution of RAG that incorporates autonomous agents to enhance retrieval, reasoning, and generation capabilities. In Agentic RAG, the retrieval path involves adaptive, goal-driven agents that dynamically explore, refine, and structure information retrieval. What is agentic RAG?
- 40. Agentic workflow User Asks Question Response Agentic Orchestration LLM News Retriever Earnings Call Retriever Summary Retriever Next Steps Retriever Customer Competitor Articles Graph Earnings Call API Customer 360 Graph Companies Graph Tools (GraphRAG) Grounding Sources
- 41. Neo4j’s Vision of Agentic Architecture Agentic Brain (KG) GenAI Ecosystem Integrations Neo4j GenAI Capabilities Vectors, LLM Callbacks Data Import Retrievers (RAG) Agentic Orchestration (Aura Service or Custom Agent Systems) Reasoning Models Memory Short Term, Long Term, Episodic, Semantic, Procedural, Reflective Neo4j agent, tools & MCP Server Planning & Reasoning Create and execute multi-step plans as a graph (store constraints rules, preferences) Agents & Tools (APIs & Databases) Decision Lineage of decision for audit and improving future decisions 1 2 3 Agents are the workers. Brain is the infrastructure that gives them memory, tools, knowledge, reasoning, and awareness.
- 42. Agentic AI in Aura Conversational Service (CCS) is an agentic, API-driven service enabling developers to build GenAI applications in ● Developer-Centric API provisioned at Database Level ● Dynamic Agent creation with configuration not code ● Simple UI to test chat experience and potentially embed in apps Coming Soon
- 43. 43 power transformation They enable & AI Knowledge graphs = data
- 45. Unlock the Power of GenAI Tim Sheedy | VP Research and Chief AI Advisor, Ecosystm
- 46. May 2025 Taking GenAI to the Next Level VP RESEARCH Tim Sheedy
- 47. 47 ecosystm.io Organisations Are on Different AI Paths “Leading businesses build cross-functional AI teams, backed by senior leadership. Collaboration sharpens business cases and directs resources where they’re needed most – especially for skills." Source: Ecosystm , 2025 3% 30% 23% 18% 25% At the planning stage Experimenting/ Creating PoCs Testing pilots Scaled deployments within specific business lines Scaled deployments across the organisation "We’ve seen digital natives do in 24 hours what takes our industry six months." “Banks modernising their cores can leap ahead, powered by embedded AI."
- 48. 48 ecosystm.io Organisations Differ in their AI Readiness Source: Ecosystm, 2025 0% 23% 4% 72% 1% Traditional Emerging Consolidating Transformative AI-First ORGANISATIONAL STRATEGY | DATA FOUNDATION | PEOPLE & SKILLS | GOVERNANCE FRAMEWORK
- 49. 49 ecosystm.io Drivers of GenAI Adoption Competitive Advantage 31% higher market differentiation. Operational Efficiency Average cost reduction of 24% in automated processes. Innovation Acceleration 41% faster product development cycles. Customer Experience 52% improvement in customer satisfaction scores. Early adopters show proven results
- 50. 50 ecosystm.io Impactful AI Use Cases Operations 63% Intelligent Document Processing 57% Payment & Invoicing Automation 52% Real-time Inventory Management IT 60% Support & Helpdesk 56% Documentation 50% Code Generation & QC Other 55% Content Strategy & Creation 55% Recruiting HR CUSTOMER, SALES & MARKETING
- 51. 51 ecosystm.io Future Plans will see Greater Adoption of AI OPERATIONS 70% Workflow Analysis 63% Fraud Detection & Prevention 61% Streamlining Risk & Compliance Processes CUSTOMER SUCCESS HR TECHNOLOGY 80% Cloud Resource Allocation & Optimisation 69% Automating Sales Processes 63% Location Based Marketing 61% Personalised Product/Service Recommendations 74% Workforce Planning 68% Talent Development & Training 62% Streamlining Employee Onboarding Source: Ecosystm, 2025 76% Network Optimisation & Performance Monitoring 70% Software Development & Testing
- 52. 52 ecosystm.io The Voice of Asia’s Leaders "Our AI-driven recruitment screening for insurance agents streamlines the selection process, quickly identifying top candidates by analysing resumes and applications.” HR LEADER "With conversational AI, we can engage customers 24/7, answering their queries and resolving issues instantly, reducing the team’s workload and enhancing CX.” CX LEADER "AI is transforming how we work – from streamlining workflows to optimising transportation routes, making operations faster and smarter.” OPERATIONS LEADER "Using AI to streamline our sales pipeline has cut down the time it takes to qualify leads, enabling our team to focus on closing more deals with greater precision.” SALES LEADER "We’re finally unlocking our data! AI agents deliver personalised customer support at scale, and AI-driven network optimisation keeps our IT running seamlessly.” DATA SCIENCE LEADER
- 53. 53 ecosystm.io Operational Disruption • System integrations break unexpectedly • Critical decision errors impact business • Productivity drops during remediation Risks of GenAI Missteps Reputational Damage • Public AI failures harm brand trust • Recovery takes 3-4x longer than building • 72% of consumers avoid brands after AI errors Financial Consequences • Failed AI projects cost $5M-$15M on average • Regulatory fines for non-compliance • Market valuation drops of 20% have been experienced
- 54. 54 ecosystm.io Barriers to GenAI Implementation Include: Cost Concerns High implementation and operational expenses Security Risks Data security and privacy vulnerabilities Regulatory Uncertainty Evolving legal frameworks across ASEAN Skills Shortage Limited AI expertise and training resources Data Challenges Poor data quality, accessibility and integration 1. 2. 3. 4. 5.
- 55. 55 ecosystm.io The Data Challenge 55 ecosystm.io Fragmented Data Landscape 86% struggle with siloed data across systems. 01 Insufficient Data Hygiene Poor data quality derails 72% of AI initiatives. 02 Contextual Limitations GenAI models often lack critical business context. 03 Linguistic Complexity Regional language diversity creates challenges. 04
- 56. 56 ecosystm.io Work is Required to Get Data “GenAI Ready” Organisational Data Readiness for GenAI (0-10) 5.3/10 COMPLIANCE 5/10 DATA SECURITY 5.3/10 DATA AVAILABILITY 5.5/10 DATA LINEAGE 7.1/10 DATA QUALITY “We actually have a semantic AI challenge: Our data has no understanding of relationships…”
- 57. GenAI LLMs overcome semantic AI challenges with billions of data points Graph Database helps you achieve the same outcome with your own organisation’s data
- 58. Our biggest GenAI challenges are not technology: Inflexible business model and processes Lack of skills Employee fear
- 59. 59 ecosystm.io Next Steps for Tech Leaders 1 Assess Readiness Evaluate current AI maturity and data quality gaps. 2 Invest in Talent Build AI skills through training and hiring. 3 Implement Semantic AI Leverage GraphDB for cohesive, contextual data use. 4 Ensure Governance Establish ethical frameworks and security controls. 59 ecosystm.io
- 60. VP Research Tim Sheedy [email protected] +61 433101233
- 62. Refresh & Recharge Next session starts at 11:10am Morning Break
- 63. AFTER LUNCH Play this brand video when session resumes
- 65. Harness Graph Power with AI Lois Ji Ronghui Data Scientist, AI Practice, GovTech Singapore (Skillsfuture Singapore Forward Deployed Team)
- 66. LOIS JI Data Scientist, AI Practice (AIP), GovTech Harness Graph Power with AI: An Insights Discovery Use Case OFFICIAL (CLOSED)
- 67. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Agenda Overview of AI & AI Reasoning How does AI reason? AI x Knowledge Graph Use Case Objective: Glean contextualized insights from unstructured data… Advantages of Graph for AI Applications Contextualized Insights Discovery with Graph Tips and Takeaways Insights delivery to End Users
- 68. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. AI Buzzwords AI Reasoning LLMs RAG Finetuning Context Window Knowledge Graph Chatbot GraphRAG Hallucination
- 69. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. How can AI perform explainable decomposition-based Reasoning? Tree-of-thought (ToT) • Multiple reasoning paths with evaluation of intermediate outputs Chain-of-thought (CoT) • Sequential task breakdown ReAct (Reason + Act) • Reasoning with tool execution Graph-of-thought (GoT) • Reusable, traceable reasoning paths, an upgrade from ToT
- 70. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. What about Agentic AI? • Systems that exhibit reasoning, planning, memory and notable degree of autonomy. • Ability to interact with its environment, learn from experience, and adapt to changes. • Capable of autonomous operation of tasks. GovTech AI Practice Agentic AI Primer: https://playbooks.aip.gov.sg/agentic-ai-primer/ Memory Memory Memory Memory Memory Proprietary Data Reasoning Graph
- 71. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. When & Why we need Knowledge Graph? • Relationship-oriented use cases. • Reasoning is more important than retrieval. • Need for Chatbot with memory.
- 72. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. CRM Insights Discovery Use Case Glean contextualised insights from large volume of unstructured data…
- 73. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Contextualized Insights with Graph ● Graph as a Bridge, a Translator ● Information Compression with “paths” ● Reasoning Augmentation for LLMs Graph is the medium, through which AI interprets and internalises tabular information.
- 74. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Solution Design: Contextualised Intelligence Using Graph Grounded Knowledge Graph with AI Models • Growing Large CRM Dataset • Informal & Inconsistent Language Use • Varying Topic Selection for Analysis • Interdependency in Issue Topics • Require Specific Enterprise Knowledge Grounding • Accuracy in Serving Non-technical End Users Need for LLMs Need for Knowledge Graph Chatbot interface for quick insights discovery Knowledge Graph traversal for causal analysis
- 75. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Preprocessing LLM Pipeline Building Graph from Unstructured Data Graph Traversal & Path Extraction Anonymization Entity Rationalization Summarisation Classification Information Extraction Clustering Aggregation Embedding Raw Data Clean Data
- 76. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Agentic Orchestration with Graph-as-an-Agent Business agents • director_agent: Proxy agent for the user, able to steer the discussion when needed • deputy_director_agent: Perform final summarisation and consolidation reporting back to user • assistant_director_agent: Gatekeeper to prevent irrelevant questions from being processed • business_analyst_agent: Break down the question into sub-queries and propose solutions Data agents • data_analyst_agent: Select and prepare python code to use tabular data for questions • python_executor_agent: Execute python code and return the outputs or any errors encountered • graph_analyst_agent: Select and prepare cypher query to use graph data for questions • cypher_executor_agent: Execute cypher query and return the outputs or any errors encountered Agentic workflow that mimics how divisions operate for data insights discovery.
- 77. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Tips: Insights delivery to End Users Hint: Users should not need to see it at all • Graphs can be overwhelming to manually traverse • Graph query is a steep learning curve for non-technical end users • Graph comprehension adds unnecessary mental load to users Instead: • Use graph as the brain to support the analytics needed. • Use graph visualisation only for investigative purposes.
- 78. COPYRIGHT OF GOVTECH © NOT TO BE REPRODUCED UNLESS WITH EXPLICIT CONSENT BY GOVTECH. Keep Calm & Graph On
- 80. Product Vision & Roadmap Michael Hunger | VP Product Innovation & Developer Strategy, Neo4j
- 81. Your business is a graph Employees Network & Security Suppliers Product Customers Transactions Process
- 82. Setting the Pace Open Source Release 2007 Cypher Query language Launch 2011 First production-ready deployment 2009 Graph Data Science and MultiDB 2020 Neo4j AuraDB Launch 2019 ISO Announces GQL Standard 2024 Native Vector Search Capabilities in Neo4j 2023 Browser and Labels 2013 OpenCypher Project Launched 2015 Distributed graph databases introduced 2017
- 84. For developers, data analysts and data scientists Premium, trusted cloud-native graph database and analytics platform Cross cloud, easy to use and enable AI accuracy
- 85. Fully Managed Seconds to Sign up Minutes to wow Days to Value Integrated Ecosystem Strategic Investments Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance c
- 86. Fully Managed Seconds to Sign up Minutes to wow Days to Value Integrated Ecosystem Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance Strategic Investments c
- 87. Unique relationship property, relationship key, property data types Quantified graph patterns Call in concurrent transactions for batch import Ultra-high speed incremental importer Incremental backup and recovery Neo4j Ops Manager Database Enhancements Graph Schema Graph Pattern Matching Offline Incremental Importer Differential Backup & Point in Time Recovery RELEASED CAPABILITIES
- 88. A superior graph native format, which groups graph data into blocks that reduces amount of data read Greater performance in memory-constrained scenarios Faster property access: properties are stored in blocks RELEASED CAPABILITIES Block Format
- 89. Change Data Capture State-of-the-art approach uses transaction log Capable of running in two modes; full or diff Integrated with Apache Kafka/Confluent RELEASED CAPABILITIES
- 90. High availability with multiDB support More features and control over the use of infrastructure resources. Scalability, allocation / reallocation, service elasticity, load balancing, automatic routing RELEASED CAPABILITIES Autonomous Clustering
- 91. Scale to 100TB+ graphs with intelligent graph sharding Separate topology and property storage for optimal performance NEW CAPABILITIES Large Graph Support Property Shards Graph Shard TXN logs Streaming WRITES Operational Analytical READS Reporting Bulk WRITES Bulk WRITES Autonomous Clustering
- 92. NEW CAPABILITIES Large Graph Support Scale to 100TB+ graphs with intelligent graph sharding Separate topology and property storage for optimal performance
- 93. Speed up analytical query up to 100x Single query is executed concurrently on multiple cores Faster insights for analytical applications and enables transactional and analytical processing in one database RELEASED CAPABILITIES Parallel Runtime
- 94. QPP can express a variable number of repetitions in a graph traversal (path) More expressive than the existing variable length paths Allow for pruning on path expansion, and as such can be order of magnitude faster than variable length paths Coming: Different MATCH modes RELEASED CAPABILITIES Quantified Path Patterns
- 95. Choose between Cypher versions: ● v1 is backwards compatible & stable ● v2 is an evolving version A version can be selected at DBMS, database and single-query level: server upgrades do not require query migration. Query Language is decoupled from the server releases. NEW CAPABILITIES Cypher Versions Cypher X Cypher Z Cypher Y Server releases (not to scale)
- 96. Conditional Queries provide more flexibility and expressive power to handle complex querying scenarios Property Access control (PBAC) data-driven rules to control READ and TRAVERSE privileges on nodes ownership, contextual access control NEW CAPABILITIES Cypher Improvements GRANT privilege property ON GRAPH {database|*} FOR ( var:label ) WHERE condition TO role GRANT privilege element ON GRAPH {database|*} FOR ()-[ var:type ]-() WHERE condition TO role MATCH (n:Person) CALL (n) { WHEN n.age > 60 THEN { SET n.ageGroup = 'Veteran' } WHEN n.age >= 35 AND n.age <= 59 THEN { SET n.ageGroup = 'Senior' } ELSE { SET n.ageGroup = 'Junior' } } RETURN n.name AS name, n.ageGroup
- 97. Strategic Investments Fully Managed GraphRAG Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance AI Accuracy Ease of Use 5Seconds to Sign up Minutes to wow Days to Value Integrated Ecosystem
- 98. Focus on Your App, Not Infrastructure! Available on Google Cloud, AWS, and Azure 99.95% uptime SLA with self-healing cluster architecture Scale up to 512GB RAM for large memory-intensive graphs Automated backups and restore Automated upgrades, zero maintenance RELEASED CAPABILITIES More than 22k Managed Databases
- 99. Comprehensive Cloud Offerings for Your Workloads Zonal (single AZ) with functionality aimed at smaller teams & departmental solutions Regional (Multi-AZ) with enterprise-grade SLA & functionality Highest tier, regional (Multi-AZ) with dedicated account provisioned per customer Perpetually free database for customers to learn and experiment with apps Now available
- 100. Get hands-on experience with AuraDB's console and tools to explore graph data Explore features before deploying workloads in production Easy transition from trial to production with ensuring continuity in your projects RELEASED CAPABILITIES AuraDB Professional Free Trial 14 day free trial available
- 101. Over 20% less expensive than previous Enterprise offering For mission-critical applications requiring enterprise-grade features Scale up to 512GB RAM for Business Critical and Virtual Dedicated Cloud with higher memory offerings coming soon AuraDB Business Critical RELEASED CAPABILITIES
- 102. Boost read performance and throughput Distribute read-heavy workloads across secondaries Horizontally scale your AuraDB instances with up to 15 read-Only secondaries Read-Only Secondaries RELEASED CAPABILITIES
- 103. Security & Compliance Trust Center SSO using IDPs such as AAD and Okta Fine-grained access control Encryption in-transit and at-rest; Customer Managed Keys Private VPC and Private Links Standard industry compliance RELEASED CAPABILITIES
- 104. Analytics Plug-in (GDS Plug-in) Run graph algorithms directly on your database cluster using a secondary node. Ideal for low-latency, lighter workloads. Works with self-managed deployments and Aura Pro. Neo4j Aura Graph Analytics Dedicated serverless offering for graph analytics that works with both Aura DB and self-managed databases. Provides better isolation, cost efficiency, and scalability but with added latency. Available through Aura. RELEASED CAPABILITIES Graph Analytics Serverless Analytics Real-time-Low Latency Scenarios Higher Latency with higher scalability & Concurrency Projection GDS Plugin AuraDB/Self Mgd Write Back Projection Write Back
- 105. Graph Analytics with Any Data Easily construct a projection from panda data frames Simplify Development Use Python to project subgraphs, run algorithms, and return results ` Ready-Made Algorithms Use out-of-the-box algorithms without coding yourself Scale Seamlessly Separate compute and storage with pay-as-you-go pricing RELEASED CAPABILITIES Neo4j Aura Graph Analytics
- 106. Enable Customers to: ● Monitor & manage all Neo4j databases across the enterprise, identify security risks ● Use Aura tooling with self managed databases ● Easily migrate self managed databases to Aura with few clicks NEW CAPABILITIES Fleet Management Aura (Unified Fleet Management) Aura Self Managed
- 107. Fully Managed GraphRAG Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance Cloud First AI Accuracy Seconds to Sign up Minutes to wow Days to Value Zero Ops Integrated Ecosystem Strategic Investments
- 108. 108 Improved productivity with GenAI copilot by helping developers write and improve Cypher queries Improved workflows with a single hub for all data management tasks Consistent user experience through a single unified console across Neo4j tools Easier collaboration as teams can share resources and collaborate on projects Secure data access with expanded roles and new access controls Seamless, Secure, GenAI-Enhanced
- 109. Unified Data Management and Visualization Now available
- 110. Automatically Import Data From Any Database Available & Adding
- 111. Schema Derived and AI-Assisted Graph Model Now available
- 112. AI-Assisted Queries and Graph Exploration
- 113. AI-Assisted, Low-Code, Interactive Dashboards Coming soon
- 114. Demo
- 115. Neo4j GraphQL library and service Deploy low code API with GraphQL Query API (Cypher over HTTPS) Allows for single Cypher requests with the response returned as JSON JDBC Driver-enable tool integrations Supports SQL translation and schema mapping RELEASED CAPABILITIES Developer APIs GraphQL Query API
- 116. 116 Powerful Graph Analytics extract features from relational data Integrate with data lakes utilize the data where it resides Graph User interface integration in other data platforms On demand compute only pay what you use Secure data access within your preferred environment Cloud native, seamless, in-place
- 117. Neo4j Graph Analytics for Snowflake GA A Native App on Snowpark Container Services Simple SQL interface for data analysts and scientists Soon 65+ GDS algorithms for ● Fraud Detection ● Customer 360 ● Recommendation ● Supply Chain This image is low res INTEGRATED DATA ECOSYSTEM Snowflake
- 118. Seamlessly uncover graph-powered insights in Snowflake AI Data Cloud 1. Project data from Snowflake tables as a graph 2. Run series of graph algorithms 3. Write results back 4. Shut down compute capacity INTEGRATED DATA ECOSYSTEM Snowflake Snowflake AI Data Cloud Snowflake Environment Pathfinding & Search Community Detection Centrality Link Prediction Similarity Graph Embeddings Graph Powered Insights–No ETL, Just SQL Graph Analytics for Snowflake
- 119. Exclusive partnership to create integrated graph analytics Deliver simplified product experience integrated with Microsoft Fabric Use your data from Microsoft OneLake INTEGRATED DATA ECOSYSTEM Microsoft Fabric GA
- 120. INTEGRATED DATA ECOSYSTEM MicrosoftFabric GA Select Lakehouse Create Graph Analyze Graph Query and explore graph results using Neo4j tools right from inside Fabric Select tables to transform data automatically into a Graph model Choose from Lakehouses in any available Workspace
- 121. Auto provision Aura Instance Select OneLake Lakehouse + Tables Auto-generated Graph Schema using LLM Import using Fabric Spark Query / Explore in Fabric directly INTEGRATED DATA ECOSYSTEM Microsoft Fabric
- 122. Strategic Investments Fully Managed Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance Cloud First Ease of Use 5Seconds to Sign up Minutes to wow Days to Value Zero Ops Integrated Ecosystem
- 123. Year of Innovation With GenAI GenAI integrations w/ Vertex AI June 2023 GenAI Stack with LangChain & Ollama Oct 2023 Vector Search & Store added to Native Graph DB August 2023 GraphRAG manifesto July 2024 Databricks Cert June 2024 Aura console with co-pilot experiences Sept 2024 Aura Pro Trial with Vector support Aug 2024 Integration with Amazon Bedrock Nov 2023 Integration with Azure OpenAI integration & Microsoft Fabric March 2024 New GraphRAG capabilities for GenAI apps April 2024 GraphRAG Package Oct 2024 Vector Optimized Aura Instances Dec 2024
- 124. GenAI Language Statistics Creativity KGs Knowledge Facts Context Knowledge Graphs Unlock GenAI Accurate Contextual Explainable
- 125. Construct knowledge graphs from unstructured data/documents using schema Implement different GraphRAG retrievers Build GenAI RAG pipelines with vector and hybrid search and GraphRAG retrieval External vector search integration GraphRAG Python Package RELEASED CAPABILITIES
- 126. Conversational Service (CCS) is an agentic, API-driven service enabling developers to build GenAI applications in minutes ● Developer-Centric API provisioned at Database Level ● Dynamic Agent creation with configuration not code ● Simple UI to test chat experience and potentially embed in apps Agentic API in Aura NEW CAPABILITIES Coming in H2
- 127. Demo
- 128. Integrated with GenAI ecosystem Orchestration & Agent Libraries Support for major frameworks like Langchain, LlamaIndex, Spring AI, Langchain4j, Haystack, Semantic Kernel, etc. Integrated with all LLM platforms like AWS, GCP, Azure, and OpenAI Coming: Model Context Protocol (MCP), CrewAI, Pydantic.AI, other Agent SDKs GenAI Ecosystem RELEASED CAPABILITIES ● Graph Connector ● CypherQAChain ● KG Construction ● Vector Index Integration ● Multiple LangChain Templates ● LangChain.js ● LangChain4j ● Cypher Data Loader ● Vector search integration ● KG construction ● Create vector index ● KG construction ● Query vector index ● Embedding Retriever ● Dynamic Document Retrieve (Cypher)
- 129. Fully Managed Seconds to Sign up Minutes to wow Days to Value Integrated Ecosystem Product Vision Summary Trusted Fundamentals Scalability with Enterprise Security, Governance, and Compliance c
- 132. Sudhir Hasbe | Chief Product Officer, Neo4j Tim Sheedy | VP Research and Chief AI Advisor, Ecosystm Lois Ji Ronghui | Data Scientist, GovTech, AI Practice Michael Hunger | VP-Product Innovation, Neo4j Moderator: Xander Smart | General Manager-ASEAN, Neo4j Q&A Panel
- 134. Sudhir Hasbe | Chief Product Officer, Neo4j Tim Sheedy | VP Research and Chief AI Advisor, Ecosystm Lois Ji Ronghui | Data Scientist, GovTech, AI Practice Michael Hunger | VP-Product Innovation, Neo4j Moderator: Xander Smart | General Manager-ASEAN, Neo4j Q&A Panel
- 135. Recharge for the Afternoon Hands on Workshops begin at 14:00 Workshop A: Create a Graph-Backed App from Scratch | Vista 2 & 3 Workshop B: Building Smarter GenAI Apps with Knowledge Graphs | Grand Ballroom Lunch Break