0% found this document useful (0 votes)

74 views

Generative AI For Media Analysis - Partner Use Case Package

Uploaded by

Félix David Mejía Mejía

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

74 views

Generative AI For Media Analysis - Partner Use Case Package

Uploaded by

Félix David Mejía Mejía

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 45

AWS Partner Network

Generative AI for Media Analysis

June 13th, 8-9 AM PT

Kevin Carlson John Rouse Amit Singh

Global Business Development Global Business Global Head, GTM & Use
Leader, Generative AI for Development Manager, Cases, Generative AI &
Media Analysis, AWS AI Services, AWS ML Partners, AWS

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
1
Generative AI for Media Analysis
Customer Success Stories

Agenda Services Overview

Demos & Reference Architectures
Ideal Customer Profile, Prospecting & Qualification
Call to Action & Partner Resources

Unlocking New Applies AI services like Amazon Rekognition

Possibilities with and Amazon Transcribe combined with
Generative AI like Amazon Bedrock and
Generative AI for Amazon Q

Media Analysis Develops deep insights into media assets

Enables new use cases for search,

monetization, video summarization, brand
safety

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
4
Find and use footage of key moments in
large video/audio archives that would
be otherwise difficult to identify

build metadata on
pre-recorded or live content to enable

Generative AI contextual advertising and binge markers

for Media
for moderation, brand
Analysis safety, brand compatibility, and
localization purposes

to summarize
content, shorten videos, upscale content,
create slow-motion effects, and more

Video Metadata Extraction Brand Safety & Contextual Ads

summarization (Audio and Visual) Fraud Detection

Sentiment Analysis Content Creation Hyper-Personalization Policy Evaluation

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
7
Generative AI for Media
Analysis Use Cases
Understanding Generative AI for Media: Customers that
have large archives of video and/or audio content, or
are routinely ingesting or streaming multimodal (audio
and video) content, can develop rich insights

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 8
Generative AI for Media
Analysis Use Cases

Generating Metadata for Archives: AI services like Transcribe,

Rekognition, and Bedrock generate metadata from archived
audio/video for enhanced search, discovery, and
monetization in the cloud

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 9
Generative AI for Media
Analysis Use Cases

Protecting Brands: Generative AI identifies brand-safe

media moments by flagging objectionable content,
violence, sensitive topics, and competing brands

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 10
Generative AI for Media
Analysis Use Cases

Understanding Context: Generative AI models

understand brands, sentiment, generate summaries,
and resolve taxonomies, enabling contextual
advertising by comprehending content context for
on-demand and live video

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 11
Generative AI for Media
Analysis Use Cases

Summarizing Media: Generative AI models leverage

transcripts to summarize content, generate insights
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. like ratings, sentiment and Ad taxonomies 12
Generative AI for Media
Analysis
Extracting insights from large audio/video
archives

Generating metadata to improve search and

monetization

Ensuring brand safety through content

moderation

Enabling contextual ad insertion through

deeper content understanding

Providing content summarization and highlight

identification
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 13
Customer Success Stories

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
14
Utilized AWS to Scale text to image to its 100 million users
Canva is an online platform for creating and editing everything
from presentations to social media posts, videos, documents,
and even websites.

Canva sets its image-creation sequence up so that after a user

enters a text prompt, it uses an Amazon SageMaker Real-Time
Inference endpoint to generate an image. When the images
are generated, the system filters them through the Amazon
Rekognition model. At the end of the pipeline, Canva displays
a selection of images to the end user. With this cutting-edge
text-to-image technology, users can create unique, high-
quality images in seconds rather than in hours or days.

“We’re not allowing users to enter prompts that could potentially

generate malicious content, and we are using Amazon Rekognition to
identify not-safe-for-work images that the model generates.”
Glen Pink, Director of ML, Canva
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Produces 100,000 Images Daily Using Generative AI on AWS

“The strategy was to use AWS services that reduce the

development workload and are simple to maintain, while
meeting low-latency and availability requirements”

- Hervé Nivon, Co-Founder and CTO, Scenario

“We are extremely bullish on this space and already expanding

to other sports properties. Fans can experience the feature
today during broadcasts of Major League Baseball (MLB), the
United States Football League (USFL) and the FIFA Women’s
World Cup™️. Catch Up With Highlights is coming to football
broadcasts this fall and more personalized content is on the
horizon. The possibilities are endless and we’re just getting
warmed up.” – Phil Martin, Vice President of Data Products

https://www.youtube.com/watch?v=QPWIrLwDZuo

https://aws.amazon.com/blogs/media/how-fox-teamed-up-with-aws-for-ai-powered-highlights/
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 17
Generative AI for Media
Analysis – Services

APPLICATIONS THAT LEVERAGE LLMs AND FMs

Amazon Q Amazon Q Amazon Q in Amazon Q in

Business Developer QuickSight Connect

Amazon Bedrock Guardrails Agents Customization Capabilities

GPUs Trainium Inferentia SageMaker

UltraClusters EFA EC2 Capacity Blocks Nitro Neuron

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved
19
Generative AI for Amazon Bedrock: multimodal embeddings

Media Analysis Amazon Rekognition: image/video analysis

Understanding Content with
Gen AI Services Amazon Transcribe: speech to text

Amazon Q: genAI–powered assistant

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
20
Amazon Choice of leading FMs through a single API

Bedrock Model customization

The easiest way to build and scale

Retrieval Augmented Generation (RAG)
generative AI applications with
foundation models (FMs)
Agents that execute multistep tasks

Security, privacy, and safety

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
21
Apply guardrails to multiple foundation
models and Agents for Amazon Bedrock
Guardrails for Configure harmful content filtering
Amazon Bedrock based on your responsible AI policies

Implement safeguards customized to your Define and disallow denied topics with
application requirements and responsible AI short natural language descriptions
policies
Redact or block sensitive information
such as PIIs, and custom Regex.

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
22
Amazon
Rekognition
Automated image, video and text
analysis

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
23
Amazon 100+ Languages and variants supported

Transcribe 10+ Set of standard features focused on

security, privacy, customization
Speech to Text
20+ Specialized features such as PII
redaction, toxicity detection and
A fully-managed and continuously generative AI-powered summarization
trained automatic speech recognition
(ASR) service built on a foundation 2 Modes: streaming, batch (2)
model

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
24
NEW Provides interactive answers, solves
problems, generates content, and takes
action

Amazon Q Understands your company information,

code, and systems
A generative AI-powered assistant Personalizes interactions based on your role
for work that is tailored to your and permissions
business
Built to be secure and private

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
26
Media2Cloud –
Metadata Extraction
• Enables content owners to generate
comprehensive knowledge graphs and metadata
profiles for their media assets, facilitating
advanced search and discovery capabilities.

• Unlock new opportunities for content

organization, repurposing, and monetization

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
27
Content Filtering Using
Custom Policy
• Implement customized policies for content
moderation and filtering based on specific
requirements.

• Facilitates comprehensive content evaluation,

ensuring compliance with regulations, brand
guidelines, or audience preferences.

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
28
Video Semantic Search
• Quick and efficient searching for specific scenes,
actions, concepts, people, or objects within large
volumes of video data using natural language
queries.

• Formulate intuitive queries and receive relevant

results, significantly enhancing the discoverability
and usability of extensive video libraries.

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
30
Ideal Customer Profile
Geography: Worldwide
Company Segment: Digital Native Business, Enterprise, Small & Medium Business (SMB), and Startup
Customer Industry: Media & Entertainment, Sports, Gaming, E-Commerce, AdTech, Social Media
Size: $100M+ in sales
Budget: $100k+

Decision Makers:
• Primary: CEOs, COO, CMO, Chief Product Officer, CTOs, Vice President, Strategy & Analytics, CRO, Trust & Safety
• Secondary: CFO, CSO, Digital Officer, Product, Engineering, Content Leaders, and Developers

Descriptors (Decision Making Factors):

• Does the customer have a large archive of media with unrealized value - either as content to improve their own
operations or for resale to others?
• Does the customer create, stream, or play media as part of their operations? Do they advertise on this media or is
there an opportunity to do so?
• Is the customer interested in moving away from targeted advertising and more towards context-aware advertising?
• Does the customer manage user-generated content, especially user-generated videos?
• Is the customer interested in new modes of search and discovery, such as visual search or using natural language to
search multimedia?
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
31
Decision Makers
Business Decision Makers Technical Decision Makers
Roles: Roles:

• Chief Operations Officer (COO) • Chief Technology Officer (CTO)

• Chief Financial Officer (CFO) • Chief Digital Officer (CDO)
• Chief Marketing Officer (CMO) • Chief Data Officer (CDO)
• Vice President Trust & Safety • Director of Product, Engineering, Content Leaders, and
Developers

Drivers
Business Decision Makers are looking to transform Drivers
their organization to meet significant business goals, Technical Decision Makers are looking for ways to reduce costs
win more customers, and beat the competition. They and increase the productivity of all systems and resources to
see generative AI to transform and overachieve their support business requirements.
business goals.
Challenges/ Pain Points
Challenges/ Pain Points
• Different teams setting up their own environments & tools
• Time/budget, difficulty scaling • Teams don’t have enough time to manage more data
• Need for insights from data and predictions
© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
32
Prospecting &
Qualification Questions

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
33
Recommended Discovery Questions
Q: Are you interested in leveraging Generative AI for content discovery, creation, conversational experiences, multimedia analysis, content moderation, or real-
time metadata generation to enhance your business operations?

Q: Does your solution provide enough flexibility to develop modern digital products according to market trends? Does your platform provide expected scalability,
reliability, and performance, especially during important events?

Q: How do you deal with manual repetitive tasks, such as analyze and search large video or audio archives? Do you see any areas for process automation in these
areas?

Q: Is it easy for your teams to find the video, audio, and image assets they need? Do you have any issues with your video teams constantly searching for content
for their projects?

Q: Do you have any specific use cases or scenarios where you would like to see the Generative AI solution in action or get a proof of concept? What problem are
you looking to solve?

Q: What types of data or content do you need to analyze or process? (e.g., text, images, videos, audio files)

Q: Are you currently evaluating or even using generative AI? Do you have generative AI resources and capabilities available in-house or would you require
support and training?

Q: What are your goals or desired outcomes from implementing Generative AI solutions? (e.g., improved content discovery, increased content creation
efficiency, enhanced customer engagement, better content moderation)

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
34
Recommended Discovery Questions
Q: What are your current challenges or pain points when it comes to:
• Analyzing and searching large video and audio archives?
• Creating real-time metadata that supports contextual advertising?
• Moderating content and keep brands safe?
• Do you operate an online platform that accepts user-generated content like social media, online forums, marketplaces etc.? This creates a need for
moderation and brand safety.
• Do you have concerns about ensuring your online content or community is safe, civil, and compliant with regulations. This suggests moderation
and brand safety is a priority.
• Do you allow your end users to upload images or videos (user generated content) to your platform or do you have to review third party content?
• How do you currently make sure that content you receive, create or that is uploaded to your platform is safe / not inappropriate, or does not cause
compliance issues?

Q: Do you have any specific requirements or constraints regarding data privacy, security, or compliance?

Q: What are your expectations regarding the accuracy, performance, and scalability of the Generative AI solution?

Q: What are your timelines or deadlines for implementing the Generative AI solution?

Q: What are your budgetary constraints or considerations for implementing and maintaining the Generative AI solution?

Q: Is there a straight forward economic advantage to using AI? Lower marketing costs or faster response times?

Q: Competitive RiskWeb
© 2024, Amazon – Are
Services,you evaluating
Inc. or its affiliates. All rights generative
reserved. AI solutions across a variety of different vendors and partners? 35
Call to Action

1 2 3 4
Review the content Develop new Use the guidance Reach out for
for this priority use solutions/offerings shared in this support on
case – ‘Generative AI on this use case or session to co-sell opportunity
for Media Analysis - infuse your existing to customers on progression, pricing,
Enhancing Search, media analysis this use case. integrations.
Monetization, and solutions with AWS
Brand Safety’. generative AI

Use case content (deck and

recording) will be released on AWS
Partner Central for on demand
access.

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved. 37
Generative AI for Media Analysis Workshop
Half-day virtual or live prescriptive engagements to help you maximize your efforts while reducing overall costs to
achieve business goals

Hands-on experience with Machine

Learning and Generative AI for media
asset enhancement

Modular approach: Explore different

media types (images, audio, video) and
use cases

Build samples and learn techniques for

media preparation, analysis, and
application development

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
38
Generative AI for Media Analysis Workshop
Consumers are relying on content not only to entertain, but also to educate, socialize, advertise and
facilitate purchasing decisions. Media is essential to how we communicate

Outcomes:

• Understand AWS Services for media analysis (images, audio, text, video)
• Learn image, text, and embeddings generation for media applications
• Gain experience with media preparation, time-based metadata, and multi-modal
metadata
• Build applications on AWS using media analysis techniques

Target Audience:

• Developers, Architects, Engineers, Data Scientists, Data Analysts, Technical Leads

GENAI - AIML | HOR | CON-SRCH-DISCO-MON

© 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.
42
Resources
Technical Guidance
• Guidance for Media2Cloud on AWS a serverless end-to-end ingest workflow to move your video assets and associated metadata to the cloud.
• Guidance for Contextual Intelligence for Advertising how to implement a contextual advertising approach with enhanced machine learning capabilities.
• Guidance for Social Media insights on AWS helps you gain insight into what your customers are saying about your products and services on social media
websites, such as Twitter.
• Guidance for Content Moderation on AWS is a serverless architecture to efficiently moderate user-contributed content

Blogs
• The next generation of media intelligence with Guidance for Media2Cloud on AWS version 4.0
• Dynamic video content moderation and policy evaluation using AWS generative AI services | Amazon Web Services
• Create super resolution for legacy media content at scale with generative AI
• Upscaling Content with Generative AI
• Moderate audio and text chats using AWS AI services and LLMs

Pricing
• Details on AWS Pricing Calculator: https://calculator.aws/#/addService/Rekognition
• Details on Rekognition Pricing can be found here: https://aws.amazon.com/rekognition/pricing/
• Details on Transcribe Pricing can be found here: https://aws.amazon.com/transcribe/pricing/
• Details on Comprehend Pricing can be found here: https://aws.amazon.com/comprehend/pricing/
• Details on Translate Pricing can be found here: https://aws.amazon.com/translate/pricing
• Details on Bedrock Pricing can be found here: https://aws.amazon.com/bedrock/pricing/
• Details on Amazon Q Pricing can be found here: https://aws.amazon.com/q/business/pricing/