Build a Real-Time Decision Support Application for Financial Market Traders with Confluent and Quix
ConfluentInc
149 views
40 slides
Oct 09, 2024
Slide 1 of 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
About This Presentation
Quix's intuitive visual programming interface and extensive library of pre-built components make it easy to build these applications without complex coding. Experience how this dynamic duo accelerates the development and deployment of your trading strategies, empowering you to make more informed...
Quix's intuitive visual programming interface and extensive library of pre-built components make it easy to build these applications without complex coding. Experience how this dynamic duo accelerates the development and deployment of your trading strategies, empowering you to make more informed decisions with real-time data!
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
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
Params,
Features
Vector Store
Telemetry
MLOps Pipelines
Model Inference
Output
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
Params,
Features
Vector Store
Telemetry
MLOps Pipelines
Model Inference
Output
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.
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
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.
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!
Confluent Perspective : AI
10
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
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
Quix Perspective : AI
13
Organizational structure determines system structure
Conway’s Law
“Organizations which design systems
are constrained to produce systems
which are copies of the
communication structures of these
organizations.”
Traditional Multi-hop and Medallion architecture
Shifting left from traditional multi-hop ETL architecture
Shift left in practice
Quix enables companies to “shift left” and empower data
teams to build with the tools they know
Build pipelines & storage to organise
data and present it to the business
Information Technology
Quix brings all your teams and data together
Data Engineering
Software Engineering Data Science
Provision & control infrastructure
and resources
Build and maintain business applications that
create and consume data
Study data to discover insights and
build ML & AI models
Reference architecture in Formula 1
Quix Streams and Quix CLI
Open source Python framework for real-time
data engineering. Streaming DataFrames
inspired by Pandas to easily move from batch
to streaming. Sources and Sinks API for
building custom integrations.
Quix
Pure Python real-time data engineering framework
Quix Cloud and Quix Edge
Unified platform for product engineering and
data teams to collaborate on building and
deploying applications and ML models
without having to manage infrastructure.
Quix Streams
Streaming DataFrames
Quix Streams
Stateful operations
Quix Streams
Stateful window operations
What’s new in Quix Streams v2.11
Sources and Sinks API CSV, InfluxDB and more)
Avro, JSON Schema, Protobuf and schema registry
support
Preview of Quix Streams v3
Branching/splitting data pipelines
Multiple topics consumed/produced in a single
application
Quix Streams roadmap
Check out the roadmap: github.com/orgs/quixio/projects/1
Contribute/open issues: github.com/quixio/quix-streams/issues
Quix platform: Quix Cloud and Quix Edge (BYOC)
Confluent Cloud native integration
?????? Flink and Quix: better together
?????? Why Flink?
●Best for data ingestion and processing of high
volume raw data
●Best for flattening, branching and joining big data
streams
●Best for software and platform engineering teams
to prepare data for downstream data analytics and
data science teams
?????? Why Quix?
●Best for building real-time operational analytics
applications
●Best for working with the data ecosystem in
Python, e.g. feature stores, vector databases, ML
model lifecycle
●Best for data analytics and data science teams to
self-serve and build applications on real-time data
streams
Solving real-time decision support for financial traders
Flink pipeline
Quix pipeline
Quix Demo
35
AWS and Brompton Bicycle
Real-time analytics for e-bike telemetry
CloudNC
Optimizing manufacturing efficiency with ML
Control
Enabling race-winning IoT connectivity with ML
Quix Templates
Pre-built projects and ready-to-run code snippets