Compose Gen-AI Apps With Real-Time Data - In Minutes, Not Weeks

As businesses strive to stay at the forefront of innovation, the ability to quickly develop scalable Generative AI (GenAI) applications is essential. Join us for an exclusive webinar featuring MIA Platform, MongoDB, and Confluent, where you'll learn how to compose GenAI apps with real-time data integration in a fraction of the time.

Discover how these three powerful platforms work together to ensure applications remain responsive, relevant, and adaptive to user preferences and contextual changes. Our experts will guide you through leveraging MIA Platform's microservices architecture and low-code development, MongoDB's flexibility, and Confluent's stream processing capabilities. Experience live demonstrations and practical insights that will transform your approach to AI-driven app development, enabling you to accelerate your development process from weeks to mere minutes. Don't miss this opportunity to keep your business at the cutting edge.

1. @yourtwitterhandle | developer.confluent.io
Starting soon…
STARTING SOOOOON..
Starting sooooon ..

2. Copyright 2021, Confluent, Inc. All rights reserved. This document may not be reproduced in any manner without the express written permission of Confluent, Inc.
Real-time AI
Model
Config
Params,
Features
Vector Store
Object store
AI-powered
Apps
Telemetry
MLOps Pipelines
Training
Data
Output

3. Goal

Partners Tech Talks are webinars where subject matter experts from a Partner talk about a
specific use case or project. The goal of Tech Talks is to provide best practices and
applications insights, along with inspiration, and help you stay up to date about innovations
in confluent ecosystem.

4. @yourtwitterhandle | developer.confluent.io
Starting soon…
STARTING SOOOOON..
Starting sooooon ..

5. The Confluent Q3 ’24 Launch

Making stream processing a developer’s best friend ????????????

STREAM
CONNECT GOVERN
Expanding our Connect with
Confluent (CwC) Partner Program

Access and configure fully managed data streams
from directly within 50+ popular data systems.
5 new integrations added in Q3 2024



!



HTTP Source & Sink Connectors v2

Two of our most popular fully managed source and
sink connectors just got even better!



Client-Side Field Level Encryption

Hey compliance teams - we don’t see your
sensitive data. Encrypt your most sensitive data
before it gets to Confluent Cloud.

PROCESS

Bringing Serverless Flink to the Tools
and Formats Developers Love



Flexible schema management- Use your own
schema formats, don’t convert stuff.

Table API (Open Preview)- Build Flink apps and
functions your way with Java, Python.

AI Model Inference (Open Preview)- Query your
AI models with Flink SQL.


More and More Private Networking

Expanding to dedicated clusters on AWS.

Confluent OEM Partner Program

Enabling MSPs, CSPs, and ISVs worldwide to provide Confluent’s data streaming platform
Easier sizing with new broker and
client metrics
Need more partitions or capacity? Our Metrics API
now includes partition warnings, CKU count, and
cluster load. We also updated the UI to show
client metrics and versions.
Confluent welcomes WarpStream!

We’re taking data streaming in your own cloud to
the next level


Build with Confluent (BwC)

11 new partners added in Q3 2024

6. Q3’ 24 Partner Program Updates
OEM Partner Program Confluent Migration Accelerator




Introducing our OEM Partner
Program!

Enabling MSPs, CSPs, and ISVs worldwide to
provide Confluent’s data streaming platform
everywhere your business is run.




Expanding our Confluent
Migration Accelerator Program!

Ensures seamless transitions from Apache
Kafka or traditional systems to Confluent,
offering exclusive resources, tailored
support, and potential funding.

Build with Confluent (BwC)

Expanding our Build with
Confluent Program!

Customers can easily jumpstart new use
cases with SIs’ data streaming solutions with
specialized software bundles, expert support,
and access to Confluent’s GTM teams.
Connect with Confluent (CwC)








Expanding our Connect with Confluent (CwC) Partner Program!

Access and configure fully managed data streams from directly within 50+ popular data systems
5 new integrations added in Q3 2024!

7. @yourtwitterhandle | developer.confluent.io
Starting soon…
STARTING SOOOOON..
Starting sooooon ..

8. Confluent Perspective : AI
8

9. How Confluent Works: Data Augmentation
Prepare data for a real-time knowledge base and contextualization in LLM queries








Schema
Registry
Confluent Cloud
Consumer
Group
Confluent-certified
Sink Connector
or Native Integration
Confluent-certified
Sink Connector
or Native Integration
Operational Data Store
(Unstructured Data)
Vector Store
Versioned Data
Products in git
Embedding API Gateway
Embedding Model Instances
Vector Embedding
Service
Schema Specs
Terraform

10. Implement Retrieval Augmented
Generation (RAG) to enrich LLMs
models with real-time updates by
using AI models as first-class
resources in Apache Flink:

●Utilize AI models directly within
Flink queries
●Manage remotely hosted AI
models with SQL DDL statements.
●Invoke remote AI model endpoints
(e.g., OpenAI, AzureML, AWS
SageMaker)
Reference a remote model from customer endpoint
Model prediction
Integrate remote AI
models seamlessly
into your data
pipeline


Preview

11. @yourtwitterhandle | developer.confluent.io
Starting soon…
STARTING SOOOOON..
Starting sooooon ..

12. https://tinyurl.com/7cr9uzh5

13. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
Mia-Platform delivers on the promise of abstracting cloud native complexity at all levels
to empower the business-IT continuum.
Mia-Platform: The Most Comprehensive IDP On The Market
⬢Founded in Milan in 2016
⬢Opened in London at the end of 2021
⬢Opened in Amsterdam in 2022
⬢Successfully launched two startups
40+Large Customers
35+Network Partners
Representative Vendor in Researches and Hype Cycles for Self Service
IDP Portal & Platform), Digital Integration Hub and Composable Technologies


Assessing Backstage for Your Internal Developer Portal - August 29th, 2024, Gary Olliffe
Develop 3 Essential Practices for Data Management Operations - July 17th, 2024, Robert Thanaraj et al.
Adopt Platform Engineering to Improve the Developer Experience - July 2nd, 2024, Gary Olliffe
Hype Cycle for Platform Engineering - June 19th, 2024, Manjunath Bhat, Bill Blosen
Hype Cycle for Application Architecture and Integration - Refreshed on July 30th, 2024, Andrew Comes, Wei Jin
Hype Cycle for Revenues and Sales Technology - June 17th, 2024, Guy Wood, Adnan Zijadic, Melissa Hilbert, Guy Wood, Ilona Hansen
Hype Cycle for Cloud Platform Services - July 24th, 2024, Robert Thanaraj, Adam Ronthal, Keith Mann, Ramke Ramakrishnan
Hype Cycle for Digital Commerce - July 7th, 2024, Sandy Shen, Tigran Egiazarov, Irina Guseva, Mike Lowndes
10
Gartner
Mentions in
2024

14. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform
Turn your Business into a Platform Company
In an evolutionary market, successful enterprises operate like tech companies.
→ An Internal Developer
Platform to pave the way
for Platform Engineering.
→ A set of Governance
tools to tackle
Application Composability

→ A Digital Integration
Hub architecture to
industrialize Data Fabric.

Mia-Platform is the unique product on the market to drive, structure
and govern all tech aspects of the Platform Company:
DATA
ENGINEERS
SOFTWARE
ENGINEERS
PLATFORM
TEAM
CLOUD
OPERATIONS & SRE

15. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
15
Infrastructure, Complex Systems, External SaaS, Cloud Services, AI Provider
DEVOPS PLATFORM
Git Pipelines Testing
Monitoring Tracing Secrets & Security
INFRASTRUCTURE PLATFORM
Execution environments: kubernetes, container managed services, functions, …
ApplicationProduct ApplicationProduct

16. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
16
Infrastructure, Complex Systems, External SaaS, Cloud Services, AI Provider
PLATFORM ORCHESTRATOR
INTERNAL
DEVELOPER PORTAL





CONSOLE IDP
CLI
DEVOPS PLATFORM
Git Pipelines Testing
Monitoring Tracing Secrets & Security
INFRASTRUCTURE PLATFORM
Execution environments: kubernetes, container managed services, functions, …
Internal Developer PlatformUI API AI
CLOUD
OPERATIONS & SRE
SOFTWARE
ENGINEERS
PLATFORM
TEAM
ApplicationProduct ApplicationProduct
Software Catalog
Scorecards
Dashboards
Health Metrics
Curated Templates
Documentation
AI Companion

17. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
17
Infrastructure, Complex Systems, External SaaS, Cloud Services, AI Provider
PLATFORM ORCHESTRATOR
INTERNAL
DEVELOPER PORTAL





CONSOLE IDP
CLI
DEVOPS PLATFORM
Git Pipelines Testing
Monitoring Tracing Secrets & Security
INFRASTRUCTURE PLATFORM
Execution environments: kubernetes, container managed services, functions, …
APPLICATION ACCESS & COMPOSABILITY
Internal Developer Platform + Composable UI API AI
CLOUD
OPERATIONS & SRE
SOFTWARE
ENGINEERS
PLATFORM
TEAM
Flow Manager
Micro Frontend Composer MFE
API Gateway API Portal Marketplace RAG Application
ApplicationProduct ApplicationProduct
Software Catalog
Scorecards
Dashboards
Health Metrics
Curated Templates
Documentation
AI Companion

18. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
18
Infrastructure, Complex Systems, External SaaS, Cloud Services, AI Provider
PLATFORM ORCHESTRATOR
INTERNAL
DEVELOPER PORTAL





CONSOLE IDP
CLI
DEVOPS PLATFORM
Git Pipelines Testing
Monitoring Tracing Secrets & Security
INFRASTRUCTURE PLATFORM
Execution environments: kubernetes, container managed services, functions, …
APPLICATION ACCESS & COMPOSABILITY
DATA PLATFORM

Fast Data CRUDData Catalog
Data Control Plane
Internal Developer Platform + Composable + DataUI API AI
CLOUD
OPERATIONS & SRE
SOFTWARE
ENGINEERS
DATA
ENGINEERS
PLATFORM
TEAM
Flow Manager
Micro Frontend Composer MFE
API Gateway API Portal Marketplace RAG Application
ApplicationProduct ApplicationProduct
Software Catalog
Scorecards
Dashboards
Health Metrics
Curated Templates
Documentation
AI Companion

19. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
Infrastructure as Code - laC tools
K8s Self-Hosted
VM  IaaS
Kubernetes - KaaS
Containers - CaaS
Application - aPaaS
Functions - fPaaS
LLM Models
Object Storage
NoSQL Data Streams
SQL SaaS
Source
Code
GitOps
Code Ship
Pipelines
Test &
Security
Manifest
Run Operate
IDE
Observe
Log/Trace
Artifactory
Repository
Secure
Business
Continuity
Connect
Auth&
Auth
Sustain.
Docs
Catalogs
Team
Third Party

20. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
Infrastructure as Code - laC tools
K8s Self-Hosted
VM  IaaS
Kubernetes - KaaS
Containers - CaaS
Application - aPaaS
Functions - fPaaS
LLM Models
Object Storage
NoSQL Data Streams
SQL SaaS
Source
Code
GitOps
Code Ship
Pipelines
Test &
Security
Manifest
Run Operate
IDE
Observe
Log/Trace
Artifactory
Repository
Secure
Business
Continuity
Connect
Auth&
Auth
Sustain.
Docs
Catalogs
Team
Internal Developer Portal API AIConsole - UI miactl - CLI
Compose
Flow
Manager
Micro Frontend
Composer MFE
API
Gateway
RAG
Application
API
Portal
Fast
Data
CRUD
SQL/NoSQL
Data
Catalog
Application As Code - Mia-Platform Orchestrator & Manifests
Mia-Platform
Third Party

21. The information included in this document is strictly confidential. Any use must be approved by Mia-Platform.
21
Mia-Platform for the Digital Integration Hub








Mainframe DBs Files MQ Cloud Applications
READ AND WRITE IN NEAR REALTIME
Pluggable Connectors
Salesforce | PostgreSQL | Golden Gate CDC | OpenLegacy | …
INGEST
Message Broker
Reliable | Scalable | Debuggable
STREAM
Raw Data Projections
Consistent | Updated | Standard
CLEANSE & VALIDATE
API & Events
Interoperable | Available 24/7 | Secure
SECURE & EXPOSE
CONFIGURE & CONTROL
⬢Microservices Creation
⬢Low-Code Configuration
⬢Multi-cloud Deployment
⬢Data Governance & Security
⬢Authentication & Authorization
⬢Monitoring & Alarms
API Events Push Notifications
API sFTP CDC Messages DB Trigger Polling JDBC
SYSTEMS OF RECORD
Core Data
Internal Apps SaaS BI ToolsLow-CodePartners
APPLICATIONS
Business Data Readers
AGGREGATE
Single Views
Unified | High-Performant | Business-ready
Mainframe DBs Files MQ Cloud Applications
Storage Bucket
Redundancy | Zero Latency | Analytics & ML
STORE
FAST DATA RUNTIME

22. How RAG Works with MongoDB?

23. RAG relies on
the ability to
search vectors

What is a vector?
A numerical representation of more
complex data, like a sentence, a
paragraph, or an image or a section of
a video.


0.30.60.80.40.2…
This vector represents the image, allowing for developers
to build systems that can, for example, find similar
images

24. How Retrieval-Augmented Generation RAG) works
extremely simplified
schema
Similarities
Relevant
Docs
Question Embeddings
How much
is the
damage
from this
photo?
Context Embeddings
A lot of
previously
evaluated
damages
Prompt AI
Starting from
that relevant
docs.

How much is
the damage
from this
photo?
Answer
The
damage is
$367

25. Combine vectors and contextual, or
operational, data to easily build elegant
Retrieval-Augmented Generation architectures
in a single developer data platform.
Operational + Vector Database
Atlas
GenAI stack requirements:
Foundation model or LLM
Application / UX
Operational Database
App Framework / Orchestration
Vector Database
MongoDB Atlas enables you to get to production fast — and stay
ahead — by minimizing cost and complexity

26. VectorOperational
Start fast and stay agile with documents that allow for easy
data modeling across multimodal data, including vectors
Simpler data modeling to start faster
Documents maps to how developers think and
code, making it easier to model relationships
between different data types, including vector
embeddings


Flexibility to evolve and stay ahead
Documents are inherently flexible and versatile for
storing richly structured multimodal data that
powers LLMs and foundation models
MongoDB Atlas
Document
Bolt-on Vector = complex data management
Relational = rigid schemas
Operational data
Vector data
Newly added data as app
requirements evolve

27. Achieve security, durability, availability and performance at
scale with multi-cloud flexibility for cloud-specific AI
tooling
Secure, available, scalable
Atlas provides mission critical security, fault tolerance,
and horizontal scaling, that satisfies even the most
demanding enterprises


Freedom to use preferred AI tech across clouds
Empower developers to use best-in-class, specialized
AI infrastructure regardless of the cloud provider




Atlas
Search node Secondary nodes Search node
Primary node
Google Vertex Amazon Bedrock Azure OpenAI

28. [0.3, 0.4, 0.2, 0.1…]
Vector
Build richer functionality with greater ease by easily
combining vector search with other types of queries
“Find me all images that are similar to the query
image and group them by photographer id”
“Find me content matching the user's query but only
content published in the last 3 months”
“Recommend restaurants that match my tastes but
only within 10 miles of my current location”
Combining with metadata for filtering
Combining with aggregations
Combining with geo-spatial search

29. How to create the RAG application from
scratch using Mia-Platform, Confluent and
MongoDB

30. Insurance
Systems
Customer
Application
FNOL System
Claims
Adjuster
Application
Insurer
Systems
CRM
Console - Internal Developer Platform
FNOLs
Policies
Customers
Claim Photos
and Evaluation
Claim Photos
Customer
Topics
Policies
Topics
FNOL
Topics
Adjuster
Topics
Claim
Customers
Topics
Embedding
Generator
Stream


Design and Runtime

Claim
Application

. Shot photo

. Ask for
evaluation


RAG
API Gateway
Single
View
Reader
BFF
OpenAI /
Gemini /
Mistral
AI
Gateway


Fast Data
Vector
Store
Claim
Single
View
Collection
with
Images
Vectorize
Single View
Creator
Embeddings
Generator

31. Live Coding