On February 7, Solvd and AWS co-hosted an exclusive lunch event titled "Your AI Problem is Really a DevOps Problem" at the AWS Santa Monica Office in California.
The event shed light on the often-overlooked yet crucial connection between AI and DevOps practices. Industry expert Luke Arrigoni, CEO of Loti, led the in-depth session. He provided the attendees with actionable insights and strategies for leveraging DevOps best practices to help them turn their ambitious AI projects into thriving real-world applications.
Key Problems Discussed
- How to reliably use AI within Media & Entertainment to make common tasks more efficient - subtitling, recognition, optimization, faster search, generating clips.
- How to streamline processes, enhance efficiency, and fine-tune the synergy between development and new AI operations?
- How to derive value from DevOps and AI investments?
- How to implement and use AI-powered tools for anomaly detection, performance monitoring, and incident response to swiftly recognize and address issues?
- How to employ AI algorithms to optimize search capabilities, enabling quick retrieval of relevant data for informed decision-making in the DevOps ecosystem?
Speakers who Attended the AWS Santa Monica Roundtable
- Luke Arrigoni, CEO of Loti shared his entrepreneurial journey and expertise in AI. He highlighted how ethical AI solutions like Loti’s facial recognition and deep fake detection are shaping the tech landscape.
- Michael Murray, Senior Solutions Architect at AWS shed light on his experience with delivering innovative cloud solutions and maximizing customer value at AWS.
- Amal De Silva, Senior Account Manager at AWS emphasized the importance of building strong customer relationships and crafting personalized solutions specifically for media and entertainment businesses.
Key Learnings to Propel Your AI Journey
- DevOps is not an afterthought. It’s an essential foundation for building and deploying robust AI solutions. Integrate DevOps principles from the outset to avoid bottlenecks and ensure smooth project execution.
- Embrace automation. Leverage tools and automation scripts to streamline repetitive tasks, freeing up resources for more strategic work. This includes automating testing, deployment, and infrastructure provisioning.
- CI/CD is your friend. Implement continuous integration and continuous delivery (CI/CD) pipelines to automate code testing, integration, and deployment. This allows for faster feedback loops and quicker iterations, leading to more robust and adaptable AI solutions.
- Think dynamically. Utilize cloud-based resources that scale up or down based on your project’s needs. This ensures efficient resource allocation and avoids unnecessary costs.
- Don’t go alone. Partner with experts like AWS, who can provide tailored solutions and support throughout your entire AI journey. Their expertise in cloud infrastructure, DevOps practices, and AI-specific tools can be invaluable.
Host's Perspective: A Resounding Success
The vibrant event for the media and entertainment industry fostered a strong sense of community, and encouraged conversations and collaboration. We noticed many commonalities amongst the participants as they have similar objectives and goals to utilize more AWS services to streamline their business processes. The conversations were engaging, with a presentation of creating an AI strategy first with a structured DevOps plan, then opened up to open discussions and architecture sessions. It is exciting to see the potential right now, as companies lean into AI. Thanks to the customers who attended, AWS and Loti. We are looking forward to more events and discussions in the future.
How to go from Model to Market
Eight things your team needs to consider for any AI project
| 1. Data Ingestion | You will need to establish ingestion contracts with data partners. Develop systems for tracking whether or not data was sent. Event tracking for when you received what you expected and so much more. |
| 2. Data Cleaning | De-duplicating and removing empty records is just a start, the real trouble starts when data needs to be normalized and fit to your enterprise data scheme. |
| 3. Terraform | The amount of infrastructure required to turn incoming data into something you can market and profit on touches outside the boundary of single server manageability and gets into git repos of infrastructure code. |
| 4. CI/CD | And all of that IaC will require unique pipelines that are raising instance types like high memory or GPU instances that your team may or may not have experience with. |
| 5. Security | Now that you have new, enriched and cleaned data. You’ll need to know how to handle this data from a security perspective. Is there PII or any other piece of data that can be dangerous? If so, who gets access and how do they get access? |
| 6. Instance Type | Will your model run best on GPU, CPU or a cloud-specific instance like Infra-2 that is optimized for inference? More expensive per hour doesn’t mean more expensive in total. Do you have the testing harness in place to evaluate the cheapest-performing solution to match your SLA? |
| 7. Serving | Will it be a batch evaluation or a real-time service? Can you run it in a python environment or does it need to run on something your team is more familiar with? Golang? C#? Understanding how to structure a runbook if there is an outage goes here too. |
| 8. Data Shift | The entire reason we are investing in AI is because we need a machine to learn and adapt to changing data. But what if data changes so much you need totally different models? How do you detect that kind of fundamental change? |
Composing with AWS Services
Useful AWS products for the media and entertainment industries
- Amazon Transcribe provides transcription services for your audio files and audio streams. It uses advanced machine learning technologies to recognize spoken words and transcribe them into text.
- Amazon Translate is a text translation service that uses advanced machine learning technologies to provide high-quality translation on demand.
- Amazon Polly is a service that turns text into lifelike speech, allowing you to create applications that talk and build entirely new categories of speech-enabled products. Amazon Polly is a Text-to-Speech service that uses advanced deep learning technologies to synthesize speech that sounds like a human voice.
- [AWS Elemental MediaConvert](lAWS Elemental MediaConvert) is a file-based video transcoding service with broadcast-grade features. It helps to create live-stream content for broadcast and multi-screen delivery at scale.
- AWS Solutions Library carries solutions built by AWS and AWS Partners for a broad range of industry and technology use cases.
- Amazon SageMaker JumpStart is a machine learning hub with foundation models, built-in algorithms, and prebuilt ML solutions that can be deployed with just a few clicks. SageMaker JumpStart gives the possibility to access pre-trained models, including foundation models, to perform tasks like article summarization and image generation.
The Road Ahead
The conversation doesn’t end here. The "Your AI Problem is Really a DevOps Problem" event ignited a spark. It’s now up to us, the AI and DevOps communities, to keep it burning bright.
- Stay curious. Explore the latest advancements in AI and DevOps, actively seeking out unconventional solutions and asking challenging questions.
- Collaborate actively. Participate in communities, attend events, and connect with fellow professionals. Share your experiences, learn from others, and foster a culture of knowledge exchange.
- Experiment fearlessly. Don't be afraid to try new things, even if they seem unconventional. Innovation thrives in an environment that encourages experimentation and calculated risks.
By embracing this collaborative spirit, we can unlock the true potential of AI and DevOps, not just for individual projects but for shaping a future where technology empowers us all.
Stay tuned and keep an eye out for upcoming events from Solvd and AWS, where you can gain further insights, connect with industry leaders, and propel your AI journey forward.
Why Choose Solvd and AWS as your Tech Partners?
Empower your AI vision with a dream team! Solvd, the AI and data specialists, combine forces with the cloud powerhouse AWS in order to provide an unbeatable partnership for your AI and DevOps journey. Leverage Solvd’s expertise in crafting ethical AI solutions and navigating the complexities of AI development. Gain access to AWS's cutting-edge cloud infrastructure, robust DevOps tools, and global reach. Together, they offer:
- Seamless integration. Solvd’s solutions are tailored to seamlessly integrate with AWS services, ensuring a smooth AI and DevOps experience.
- End-to-end support. From ideation to deployment, Solvd and AWS guide you every step of the way, providing expert advice and technical assistance.
- Scalability and agility. Scale your AI initiatives effortlessly with AWS’s infrastructure, adapting to your evolving needs with agility.
- Innovation at your fingertips. Tap into the combined expertise of Solvd and AWS to explore cutting-edge AI and DevOps advancements.
Don’t embark on this transformative journey alone. Choose Solvd and AWS as your trusted companions, and turn your AI vision into a thriving reality.
