Product Strategy & Roadmap
Anurag Tandon
VP Product Management, Neo4j
The world is changing
As things get more connected, data does too.
Neo4j Inc. All rights reserved 2023
2
The world is changing
Organizations need to transform from data
to insights to knowledge
Neo4j Inc. All rights reserved 2023
3
KNOWLEDGE GRAPHS
Enable various use cases across enterprises
Transactions
graph
Consumer
graph
Parts
graph
Digital Twin
graph
Neo4j Inc. All rights reserved 2023
4
Top 3 trends impacting developers
5
Market Trends:
1. Cloud (multi) adoption is accelerating
2. Most use cases blend transactional and
analytical capabilities
3. Enterprise interest in LLM/GenAI is
through the roof but adoption is limited
due to hallucination
Product Vision
Cloud First With Ecosystem Focus
Cloud Scale with Enterprise Security, Governance & Compliance
Trusted Fundamentals
Most advanced Graph Algorithms
with in memory projections
Enable intelligent GenAI apps with
Knowledge graphs + LLMs
Integrate with GenAI platforms
Cloud native service integrations to
simplify application development
Fully Managed Database for 0
operations experience so developers can
focus on building applications
Get started with Free Database
Self-service capabilities that enable
developers to improve time to value
Simplify APIs and tools to improve
productivity of developers
Enhanced developer tooling
Cloud First
HOW
WE
ARE
EXECUTING
seconds to sign up
minutes to wow/data
days to value
5
Premium & trusted cloud-native graph database and analytics platform that is cross cloud,
easy to use & AI Enabler for developers, data analysts and data scientists
AI Enabler
Ease of Use
Neo4j Inc. All rights reserved 2023
6
Neo4j Database
Trusted Fundamentals
Neo4j Inc. All rights reserved 2023
7
Graph Schema: New constraints on nodes, relationships and
properties: Unique relationship property, Relationship key,
Property data types
Graph Pattern Matching: Improved expressivity of graph
navigation with quantified graph patterns. More powerful and
performant syntax to navigate and traverse your graph.
Call in Transactions: for batched import, control transaction
sizes from Cypher.
Incremental Importer: Ultra-high speed method of loading data
incrementally (10-100x faster than transactional).
Backup & Recovery: Existing full backups can be updated with
differential data instead of recreating a new full backup, saving
a large amount of storage. Administrators can now execute
point-in-time restore. API to ease operability
Neo4j Ops Manager: To monitor Neo4j deployments
Neo4j 5.x new capabilities
Database Enhancements GRAPH PATTERN MATCHING
GRAPH SCHEMA
DIFFERENTIAL BACKUP &
POINT IN TIME RECOVERY
OFFLINE INCREMENTAL IMPORTER
Neo4j Inc. All rights reserved 2023
9
Autonomous Clustering
Easy, Automated Horizontal Scale-Out
Fabric
Federated Queries and Sharded Graphs
A significant set of improvements in Neo4j clustering architecture, with
more features and control over the use of infrastructure resources.
Scalability, allocation / reallocation, service elasticity, load balancing,
automatic routing are some of the main features added to the existing
architecture.
Query multiple business graphs
Chain queries for sophisticated real-time analysis
Hybrid cloud queries
Neo4j 5.x new capabilities
Unbounded Scalability To Harness All Data
1. Admin specifies the
databases, primaries for Fault
Tolerance and secondaries for
read scalability
2. Autonomous Cluster allocates
them to suitable servers
3. Admin can add or remove
servers, and then
reallocate the databases across
the cluster
– Leverage customer hints to create sharding
strategy
– Enable creation of schema/metadata for sharding
– Automatically increase shards based on customer
configuration
Mapping
Sharding
Rules
– Create Fabric planner to push data and queries to
right shards
– Overtime support distributed transactions &
proxy nodes
– Make it available in Aura
Neo4j Inc. All rights reserved 2023
10
Coming 2024+
Neo4j new capabilities
Fabric Auto Sharding
Challenge: Analytical queries are graph global,
not anchored to particular nodes, and traversing
many parts of the graph. Analytical Queries can
benefit from multi threading.
Solution: With Parallel Runtime, a single query is
executed concurrently on multiple cores,
providing a better option for a significant number
of use cases.
Benefits: Faster insights for analytical applications
& enables transactional and analytical processing in
one database
Neo4j 5.x new capabilities
Parallel Runtime: Speed up analytical query up to 100x
Announcing
Parallel Runtime Speedup
Up to 100x faster analytical queries by adding CPU cores
Neo4j Inc. All rights reserved 2023
12
Analytical query from StackOverflow dataset:
For all questions for n months from x date, return
● the tags
● the number of distinct users posting the
questions
● the maximum score for the questions
● for all answered questions, the average
number of answers
● the number of distinct questions
Challenge: Need database change tracking to enable swift
mission-critical actions, cloud database integrations, avoid
disruption to production system, and ensure data traceability.
Solution:
● State-of-the-art approach uses transaction log based
Change Data Capture (CDC),
● Capable of running in two modes; full where all information
on a node or relationship is sent if there is a change, or diff
mode where only the information that changed is sent,
● A Cypher command that enables CDC on a named
database. A procedural API that can be called by partner
applications or those developed by customers
● Used by Neo4j Connector for Apache Kafka/Confluent.
Benefits: Confidence in Neo4j as db system of record, Real-time
streaming to target systems to support event-driven mission
critical decisions, Audit and compliance
Announcing
Neo4j 5.x new capabilities
Change Data Capture: Automated real-time change tracking
Cloud First
Neo4j Inc. All rights reserved 2023
14
FAST FLEXIBLE RELIABLE EASY
● Automated Upgrades, Maintenance
● Scalable and Elastic, On-Demand
● Enterprise-Grade Security
● High Availability
● Simple Pricing, Consumption-Based
● Procure through Aura Console or via
Cloud Marketplace
Neo4j Aura
Fully Managed Cloud Service on all clouds
● Ubiquitous availability of Aura in all major
clouds: GCP, AWS, Azure
● Enterprise-ready Aura
﹣ SOC II Type 2 compliance
﹣ Better DevOps with AuraDB APIs
﹣ Easier RBAC configuration via Aura
console
﹣ Better observability with security log
forwarding (EAP) and Performance
metrics forwarding (EAP)
﹣ Private Link
﹣ CMEK-Coming soon
Neo4j Aura
2023 Key Capabilities
Neo4j Inc. All rights reserved 2023
17
Cloud Data Ecosystem
Plugs Into Your Existing Ecosystem
Self Service (Ease of Use)
Neo4j Inc. All rights reserved 2023
18
Self Service Tooling And Developer Experience
Comprehensive set of tools for self-service
ops manager
data importer
Visualize and explore your data
Query editor and results visualizer
Code-free data loader and modeler
NeoDash (BI)
Neo4j Inc. All rights reserved 2023
19
Self Service Enhancements
1. Data Import: Model & Load Your First Graph:
Neo4j Importer
2. Bloom: Explore Graph Algorithms, Time Slicer,
Cypher Actions, Search improvements
3. Browser/Query-Better favorites, history +
visualization
4. Unified Developer Experience with Neo4j
Workspace
Roadmap (2024)
1. Simplified import from various relational and
cloud systems
2. New Graph Visualization Library
3. Improved Cypher Development Support-
VS Code Extension
Self Service Tooling And Developer Experience
Comprehensive set of tools for self-service
Self Service Tooling And Developer Experience
Better client surface for Developers and Data Scientists
Neo4j GraphQL Library
● Build low code API with GraphQL Library & Toolbox
● GraphQL support has enabled 1B queries in Aura
Simplified Drivers API for Neo4j
● Simplified API Driver object - driver.executeQuery() -
returns results directly into native formats
● Driver APIs automates various capabilities like Sessions,
Transaction Functions, Bookmarks
GDS Native Python client
● Wraps the Neo4j python driver (dataframe support)
● Run GDS algorithms just like you would any python
function
● run_cypher lets you execute Cypher statements
● Pythonic features: support for graph and model objects
AI Enabler
Graph Data Science & Generative AI
Neo4j Inc. All rights reserved 2023
22
Neo4j Inc. All rights reserved 2023
23
Graph Data Science
Make Sense Of Data Relationships
Machine Learning Pipeline
Pl
ay
s
Lives_in
In_sport
Likes
F
a
n
_
o
f
Plays_for
K
n
o
w
s
Knows
Knows
K
n
o
w
s
Explore the hidden patterns and features in your data
What’s important? What’s unusual? What’s next?
Neo4j Inc. All rights reserved 2023
24
Graph Data Science
Make Sense Of Data Relationships
Over 65 efficient, parallelized algorithms. Iterate fast with different data sets & models,
version trained models.
Bring the context of your connected data into
a format that other pipelines can ingest.
The Largest Catalog of
Graph Algorithms
Native Graph Catalog and
Analytics Workspace
Graph Vector Embeddings
for Machine Learning
Topological Sort Algorithm
Identify dependencies
Longest Path Algorithm
Identify critical paths
Knowledge Graph
Embeddings
Discover missing relationships
Neo4j Inc. All rights reserved 2023
25
Key Use Cases:
GenAI, Semantic Search Recommendations,
Life Sciences
Key Use Cases:
Supply Chain and Network Routing
Key Use Cases:
Supply Chain, Inventory Management,
Resource Allocation, and Build Management
Announcing
Graph Data Science
What’s New in Graph Data Science Library
Step 1: Build a graph Step 2: Export your graph using
the Python client
Step 3: Train KGE model
with PyKeen, DGL, Pytorch
Geometric, etc.
Step 4: Import embeddings
created from KGE training as
properties in your graph
Step 5: Create candidate
relationships of possible
connections using KNN
Step 6: Check node pairs
for missing connections
Announcing
Graph Data Science
Discover Missing Relationships with KGE
AI Enabler
Graph Data Science & Generative AI
Neo4j Inc. All rights reserved 2023
27
Lack of enterprise domain knowledge
Limited input sizes for fine tuning
Inability to verify answers
Sensitive to prompt phrasing & injection
Hallucinates
ETHICAL & DATA BIAS CONCERNS
28 Neo4j Inc. All rights reserved 2023
Powering Generative AI Apps
What GenAI can’t do!
29
Knowledge
Facts
Context
Language
Statistics
Creativity
KGs LLMs
+
BETTER TOGETHER
Generative AI, LLMs and Knowledge Graphs
Growing conviction that not only will the adoption of GenAI and LLMs grow
substantially but also that we have an essential role to play
Neo4j Inc. All rights reserved 2023
Neo4j Inc. All rights reserved 2023
30
Powering Generative AI Apps
Better Pipeline With LLMs + Neo4j
Neo4j Inc. All rights reserved 2023
New Capabilities
● Vector Search & Storage
-Enabling RAG (Retrieval Augmented
Generation)
● Cypher: Real Time integration with
Embedding APIs & LLM Models
● Integrations with LLM Platforms: OpenAI +
MS Azure OpenAI, VertexAI, AWS Bedrock
● Integration with GenAI Frameworks:
LangChain, LlamaIndex
● GenAI Dev Stack: Docker + Neo4j +
Langchain + Ollama
Roadmap (2024+)
● Co-Pilot in Browser for autocomplete &
cypher generation
● Bloom & NeoDash NL Integration
Powering Generative AI Apps
Neo4j GenAI Innovations and Roadmap
Recap
New Announcements
Neo4j Inc. All rights reserved 2023
32
Product Roadmap Recap
Driving Innovation with Neo4j for Developers
Knowledge
Graph Embeddings (NEW)
Longest Path Algorithm (NEW)
Topological Sort Algorithm(NEW)
Aura Enterprise Database on
all clouds (AWS, GCP, Azure)
SOC II Type 2 compliance,
AuraDB APIs, RBAC
configuration
Better observability with
security log forwarding
Performance metrics
forwarding (EAP)
Private Link & CMEK (Coming
soon)
Unified Developer
Experience with
Workspace
Self-service Data Import
GraphQL Support
Simplified Drivers API
Bloom support for GDS
algorithms
GDS Python API
Cloud First AI Enabler
Ease of Use
Neo4j Inc. All rights reserved 2023
33
Parallel Runtime for faster
analytical Queries (NEW)
Change Data Capture better
data integration (NEW)
Autonomous clustering and
Fabric for limitless scalability
Graph Schema, Improved
Backup recovery, incremental
import
Neo4j Ops Manager for
managing databases
Trusted
Fundamentals
Vector Search (NEW)
Real Time integration with
Embedding APIs & LLM Models
Integrations with OpenAI + MS
Azure OpenAI, VertexAI, AWS
Bedrock, Langchain, LlamaIndex
GDS
GenAI
Thank You
Anurag Tandon
anurag.tandon@neo4j.com

Peek into Neo4j Product Strategy and Roadmap

  • 1.
    Product Strategy &Roadmap Anurag Tandon VP Product Management, Neo4j
  • 2.
    The world ischanging As things get more connected, data does too. Neo4j Inc. All rights reserved 2023 2
  • 3.
    The world ischanging Organizations need to transform from data to insights to knowledge Neo4j Inc. All rights reserved 2023 3
  • 4.
    KNOWLEDGE GRAPHS Enable varioususe cases across enterprises Transactions graph Consumer graph Parts graph Digital Twin graph Neo4j Inc. All rights reserved 2023 4
  • 5.
    Top 3 trendsimpacting developers 5 Market Trends: 1. Cloud (multi) adoption is accelerating 2. Most use cases blend transactional and analytical capabilities 3. Enterprise interest in LLM/GenAI is through the roof but adoption is limited due to hallucination
  • 6.
    Product Vision Cloud FirstWith Ecosystem Focus Cloud Scale with Enterprise Security, Governance & Compliance Trusted Fundamentals Most advanced Graph Algorithms with in memory projections Enable intelligent GenAI apps with Knowledge graphs + LLMs Integrate with GenAI platforms Cloud native service integrations to simplify application development Fully Managed Database for 0 operations experience so developers can focus on building applications Get started with Free Database Self-service capabilities that enable developers to improve time to value Simplify APIs and tools to improve productivity of developers Enhanced developer tooling Cloud First HOW WE ARE EXECUTING seconds to sign up minutes to wow/data days to value 5 Premium & trusted cloud-native graph database and analytics platform that is cross cloud, easy to use & AI Enabler for developers, data analysts and data scientists AI Enabler Ease of Use Neo4j Inc. All rights reserved 2023 6
  • 7.
    Neo4j Database Trusted Fundamentals Neo4jInc. All rights reserved 2023 7
  • 8.
    Graph Schema: Newconstraints on nodes, relationships and properties: Unique relationship property, Relationship key, Property data types Graph Pattern Matching: Improved expressivity of graph navigation with quantified graph patterns. More powerful and performant syntax to navigate and traverse your graph. Call in Transactions: for batched import, control transaction sizes from Cypher. Incremental Importer: Ultra-high speed method of loading data incrementally (10-100x faster than transactional). Backup & Recovery: Existing full backups can be updated with differential data instead of recreating a new full backup, saving a large amount of storage. Administrators can now execute point-in-time restore. API to ease operability Neo4j Ops Manager: To monitor Neo4j deployments Neo4j 5.x new capabilities Database Enhancements GRAPH PATTERN MATCHING GRAPH SCHEMA DIFFERENTIAL BACKUP & POINT IN TIME RECOVERY OFFLINE INCREMENTAL IMPORTER
  • 9.
    Neo4j Inc. Allrights reserved 2023 9 Autonomous Clustering Easy, Automated Horizontal Scale-Out Fabric Federated Queries and Sharded Graphs A significant set of improvements in Neo4j clustering architecture, with more features and control over the use of infrastructure resources. Scalability, allocation / reallocation, service elasticity, load balancing, automatic routing are some of the main features added to the existing architecture. Query multiple business graphs Chain queries for sophisticated real-time analysis Hybrid cloud queries Neo4j 5.x new capabilities Unbounded Scalability To Harness All Data 1. Admin specifies the databases, primaries for Fault Tolerance and secondaries for read scalability 2. Autonomous Cluster allocates them to suitable servers 3. Admin can add or remove servers, and then reallocate the databases across the cluster
  • 10.
    – Leverage customerhints to create sharding strategy – Enable creation of schema/metadata for sharding – Automatically increase shards based on customer configuration Mapping Sharding Rules – Create Fabric planner to push data and queries to right shards – Overtime support distributed transactions & proxy nodes – Make it available in Aura Neo4j Inc. All rights reserved 2023 10 Coming 2024+ Neo4j new capabilities Fabric Auto Sharding
  • 11.
    Challenge: Analytical queriesare graph global, not anchored to particular nodes, and traversing many parts of the graph. Analytical Queries can benefit from multi threading. Solution: With Parallel Runtime, a single query is executed concurrently on multiple cores, providing a better option for a significant number of use cases. Benefits: Faster insights for analytical applications & enables transactional and analytical processing in one database Neo4j 5.x new capabilities Parallel Runtime: Speed up analytical query up to 100x Announcing
  • 12.
    Parallel Runtime Speedup Upto 100x faster analytical queries by adding CPU cores Neo4j Inc. All rights reserved 2023 12 Analytical query from StackOverflow dataset: For all questions for n months from x date, return ● the tags ● the number of distinct users posting the questions ● the maximum score for the questions ● for all answered questions, the average number of answers ● the number of distinct questions
  • 13.
    Challenge: Need databasechange tracking to enable swift mission-critical actions, cloud database integrations, avoid disruption to production system, and ensure data traceability. Solution: ● State-of-the-art approach uses transaction log based Change Data Capture (CDC), ● Capable of running in two modes; full where all information on a node or relationship is sent if there is a change, or diff mode where only the information that changed is sent, ● A Cypher command that enables CDC on a named database. A procedural API that can be called by partner applications or those developed by customers ● Used by Neo4j Connector for Apache Kafka/Confluent. Benefits: Confidence in Neo4j as db system of record, Real-time streaming to target systems to support event-driven mission critical decisions, Audit and compliance Announcing Neo4j 5.x new capabilities Change Data Capture: Automated real-time change tracking
  • 14.
    Cloud First Neo4j Inc.All rights reserved 2023 14
  • 15.
    FAST FLEXIBLE RELIABLEEASY ● Automated Upgrades, Maintenance ● Scalable and Elastic, On-Demand ● Enterprise-Grade Security ● High Availability ● Simple Pricing, Consumption-Based ● Procure through Aura Console or via Cloud Marketplace Neo4j Aura Fully Managed Cloud Service on all clouds
  • 16.
    ● Ubiquitous availabilityof Aura in all major clouds: GCP, AWS, Azure ● Enterprise-ready Aura ﹣ SOC II Type 2 compliance ﹣ Better DevOps with AuraDB APIs ﹣ Easier RBAC configuration via Aura console ﹣ Better observability with security log forwarding (EAP) and Performance metrics forwarding (EAP) ﹣ Private Link ﹣ CMEK-Coming soon Neo4j Aura 2023 Key Capabilities
  • 17.
    Neo4j Inc. Allrights reserved 2023 17 Cloud Data Ecosystem Plugs Into Your Existing Ecosystem
  • 18.
    Self Service (Easeof Use) Neo4j Inc. All rights reserved 2023 18
  • 19.
    Self Service ToolingAnd Developer Experience Comprehensive set of tools for self-service ops manager data importer Visualize and explore your data Query editor and results visualizer Code-free data loader and modeler NeoDash (BI) Neo4j Inc. All rights reserved 2023 19
  • 20.
    Self Service Enhancements 1.Data Import: Model & Load Your First Graph: Neo4j Importer 2. Bloom: Explore Graph Algorithms, Time Slicer, Cypher Actions, Search improvements 3. Browser/Query-Better favorites, history + visualization 4. Unified Developer Experience with Neo4j Workspace Roadmap (2024) 1. Simplified import from various relational and cloud systems 2. New Graph Visualization Library 3. Improved Cypher Development Support- VS Code Extension Self Service Tooling And Developer Experience Comprehensive set of tools for self-service
  • 21.
    Self Service ToolingAnd Developer Experience Better client surface for Developers and Data Scientists Neo4j GraphQL Library ● Build low code API with GraphQL Library & Toolbox ● GraphQL support has enabled 1B queries in Aura Simplified Drivers API for Neo4j ● Simplified API Driver object - driver.executeQuery() - returns results directly into native formats ● Driver APIs automates various capabilities like Sessions, Transaction Functions, Bookmarks GDS Native Python client ● Wraps the Neo4j python driver (dataframe support) ● Run GDS algorithms just like you would any python function ● run_cypher lets you execute Cypher statements ● Pythonic features: support for graph and model objects
  • 22.
    AI Enabler Graph DataScience & Generative AI Neo4j Inc. All rights reserved 2023 22
  • 23.
    Neo4j Inc. Allrights reserved 2023 23 Graph Data Science Make Sense Of Data Relationships Machine Learning Pipeline Pl ay s Lives_in In_sport Likes F a n _ o f Plays_for K n o w s Knows Knows K n o w s Explore the hidden patterns and features in your data What’s important? What’s unusual? What’s next?
  • 24.
    Neo4j Inc. Allrights reserved 2023 24 Graph Data Science Make Sense Of Data Relationships Over 65 efficient, parallelized algorithms. Iterate fast with different data sets & models, version trained models. Bring the context of your connected data into a format that other pipelines can ingest. The Largest Catalog of Graph Algorithms Native Graph Catalog and Analytics Workspace Graph Vector Embeddings for Machine Learning
  • 25.
    Topological Sort Algorithm Identifydependencies Longest Path Algorithm Identify critical paths Knowledge Graph Embeddings Discover missing relationships Neo4j Inc. All rights reserved 2023 25 Key Use Cases: GenAI, Semantic Search Recommendations, Life Sciences Key Use Cases: Supply Chain and Network Routing Key Use Cases: Supply Chain, Inventory Management, Resource Allocation, and Build Management Announcing Graph Data Science What’s New in Graph Data Science Library
  • 26.
    Step 1: Builda graph Step 2: Export your graph using the Python client Step 3: Train KGE model with PyKeen, DGL, Pytorch Geometric, etc. Step 4: Import embeddings created from KGE training as properties in your graph Step 5: Create candidate relationships of possible connections using KNN Step 6: Check node pairs for missing connections Announcing Graph Data Science Discover Missing Relationships with KGE
  • 27.
    AI Enabler Graph DataScience & Generative AI Neo4j Inc. All rights reserved 2023 27
  • 28.
    Lack of enterprisedomain knowledge Limited input sizes for fine tuning Inability to verify answers Sensitive to prompt phrasing & injection Hallucinates ETHICAL & DATA BIAS CONCERNS 28 Neo4j Inc. All rights reserved 2023 Powering Generative AI Apps What GenAI can’t do!
  • 29.
    29 Knowledge Facts Context Language Statistics Creativity KGs LLMs + BETTER TOGETHER GenerativeAI, LLMs and Knowledge Graphs Growing conviction that not only will the adoption of GenAI and LLMs grow substantially but also that we have an essential role to play Neo4j Inc. All rights reserved 2023
  • 30.
    Neo4j Inc. Allrights reserved 2023 30 Powering Generative AI Apps Better Pipeline With LLMs + Neo4j
  • 31.
    Neo4j Inc. Allrights reserved 2023 New Capabilities ● Vector Search & Storage -Enabling RAG (Retrieval Augmented Generation) ● Cypher: Real Time integration with Embedding APIs & LLM Models ● Integrations with LLM Platforms: OpenAI + MS Azure OpenAI, VertexAI, AWS Bedrock ● Integration with GenAI Frameworks: LangChain, LlamaIndex ● GenAI Dev Stack: Docker + Neo4j + Langchain + Ollama Roadmap (2024+) ● Co-Pilot in Browser for autocomplete & cypher generation ● Bloom & NeoDash NL Integration Powering Generative AI Apps Neo4j GenAI Innovations and Roadmap
  • 32.
    Recap New Announcements Neo4j Inc.All rights reserved 2023 32
  • 33.
    Product Roadmap Recap DrivingInnovation with Neo4j for Developers Knowledge Graph Embeddings (NEW) Longest Path Algorithm (NEW) Topological Sort Algorithm(NEW) Aura Enterprise Database on all clouds (AWS, GCP, Azure) SOC II Type 2 compliance, AuraDB APIs, RBAC configuration Better observability with security log forwarding Performance metrics forwarding (EAP) Private Link & CMEK (Coming soon) Unified Developer Experience with Workspace Self-service Data Import GraphQL Support Simplified Drivers API Bloom support for GDS algorithms GDS Python API Cloud First AI Enabler Ease of Use Neo4j Inc. All rights reserved 2023 33 Parallel Runtime for faster analytical Queries (NEW) Change Data Capture better data integration (NEW) Autonomous clustering and Fabric for limitless scalability Graph Schema, Improved Backup recovery, incremental import Neo4j Ops Manager for managing databases Trusted Fundamentals Vector Search (NEW) Real Time integration with Embedding APIs & LLM Models Integrations with OpenAI + MS Azure OpenAI, VertexAI, AWS Bedrock, Langchain, LlamaIndex GDS GenAI
  • 34.