Fire up the opportunities with GenAI on AWS
Avahi’s Ignition AI package offerings on AWS
Our customizable AI-ML solutions and services can engage the customer at whatever stage of adoption:
- Ignition AI Workshop - free half-day seminar designed to uncover an initial use case and establish a roadmap to adoption.
- Ignition AI Explorer - AWS Managed Services: Bedrock, Titan, CodeWhisperer to deliver a single use case PoC, typically a four week engagement.
- Ignition AI Jumpstart - SageMaker Jumpstart to test and deploy a variety of FMs to deliver MVP introducing customer data and MLOps with SageMaker; typically eight to twelve week engagement.
- Ignition CodeWhisperer - Avahi helps direct a data-driven PoC to help customers and developers alike achieve efficiencies with code development through the use of generative AI - 30-day timeline.
- Ignition AI at-Scale - a production-grade solution capable of delivering model inference at-scale, fully integrated MLOps with SageMaker and data curation and governance. Custom scope of work and timelines vary based on complexity.
Avahi's Ignition AI Workshop on AWS
Avahi offers a free half-day seminar to help kickstart your journey with AI on AWS. Attendees learn about the foundations of generative AI, large language models (LLMs) through today’s foundational models (FM), the technologies behind the rapid mass adoption of generative AI, and, importantly, the ethical considerations of this technology.
At its conclusion, attendees have the baseline knowledge to identify potential use cases, be introduced to AWS GenAI core services, and know how you can leverage AWS and Avahi to guide your journey.
Attendees participate:
- Amazon Bedrock
- Amazon Titan
- CodeWhisperer
- SageMaker Jumpstart
- Trainium
- Inferentia2
Ignition AI Workshop | Deliverables
- Insights into the evolution of generative AI from massive amounts of data collected and used to train Large Language Models (LLM) through today’s Foundational Models.
- Ethical considerations.
- How Foundational Models are transforming business, organizations, and government.
- Understanding of the core technologies driving generative AI’s rapid adoption.
- The AWS services helping to drive the democratization of GenAI.
- Amazon Bedrock, Titan, CodeWhisperer, and Kendra.
- SageMaker Jumpstart, Trainium and Inferentia2.
- Identify use case(s) and demonstrate the AWS Services to help achieve early success.
- A clear path forward to a PoC engagement.
Avahi's Ignition AI Explorer | PoC engagement
Core Technologies:
- Amazon Bedrock
- Amazon Titan
- Stability.ai (open source)
- AI21
- CodeWhisperer
Ignition AI Explorer | Deliverables
- Ignition Workshop.
- Align customer’s business requirements with use case.
- QuickStart deployment for a single use case.
- Choose an open-source FM from a qualified list.
- Implement AWS best practice guardrails.
- Foundation for continued exploration and adoption.
- Engagement timeline is approximately four weeks
Exploring Use Cases
- Text Generation: Develop original content from stories to essays.
- Chatbots: Enhance customer interactions with AI-driven interfaces.
- Search: Efficiently extract and synthesize information.
- Text Summarization: Generate concise versions of extensive textual content.
- Image Generation: Produce images from textual prompts.
- Image Classification: Automate image categorization and captioning.
Ignition AI Jumpstart | MVP engagement
- SageMaker
- Amazon Bedrock
- Amazon Titan
- CodeWhisperer
- Kendra
- Stability.ai (open source)
- AI21
- AWS services used to gather, transform, store and govern customer data
Ignition AI Jumpstart | Deliverables
- The Ignition AI Workshop.
- Helping the customer select the right Foundational Model (FM).
- QuickStart deployment of AWS best practices and guardrails.
- Gather, curate and governance of customer data.
- Model training, MLOps w/ SageMaker pipeline.
- MVP deployment with model monitoring.
- Best practice recommendations for deployment at-scale.
- Engagement timeline is based on specific requirements, typically eight-twelve weeks.
Exploring Use Cases
- Text Generation: Develop original content from stories to essays.
- Chatbots: Enhance customer interactions with AI-driven interfaces.
- Search: Efficiently extract and synthesize information.
- Text Summarization: Generate concise versions of extensive textual content.
- Image Generation: Produce images from textual prompts.
- Image Classification: Automate image categorization and captioning.
Ignition AI AWS CodeWhisperer | PoC Engagement
Avahi working with the AWS CodeWhisperer team to provide a turnkey PoC for customers to evaluate the efficacy of the service. Using pre-established quickstart deployments, prescriptive KPIs and measurements, and visualizations – allow stakeholders to make data-driven (quantitative) and developer feedback (qualitative) decisions.The timeline is typically 30-days and ideally between 15 – 25 developers, the number of developers can vary significantly and still provide value for the customer.
Core Technologies Focus:
- AWS CodeWhisperer
- Amazon SageMaker Studio, Cloud9 and supported IDEs
- CodePipeline, CodeBuild, CodeDeploy
- Supported programming languages
FAQs
Let us share our knowledge, experience and insights from building 100s of solutions across many industries as a cloud-first AWS Consulting Partner.
- AI-ML at-scale, generative AI models, tools and AWS Managed Services
- Data, Analytics
- Database Modernization
- Migrations: on-perm to AWS | cloud-to-cloud
- DevOps
- AWS MAP and other AWS enablement programs
- Well-Architected
- We work closely with the Stability.ai integration team on several projects and PoCs.
- Migration of FM model inference-only to AWS.
- Containerization of FM using Amazon ECS.
- Private hosting of the Stability FM in AWS for a HiTrust environment.
- Delivered a successful PoC using Amazon HealthScribe, comparing its’ viability with traditional Amazon Transcribe.
- Working w/ AWS CodeWhisperer team to deliver customized workshops and PoCs.
- Ignite AI Workshop
- Ignite Explorer AI PoC
- Ignite AWS CodeWhisperer
- Ignite Jumpstart Engagement MVP
Customer | Avahi | Outcome |
---|---|---|
We’re focused on current initiatives but would like to explore its possibilities… | Avahi, can set up a call to address your questions and explain our services… | Ignite AI Workshop 1/2 day |
We’re evaluating specific use cases but need more help choosing the right models, AWS services and tools… | Avahi has the experience and knowledge to help you avoid common mistakes and select the FMs and services for your use case and budget… | Ignite AI Explorer typically 4-weeks |
Our company is currently running a PoC but needs to understand how to securely integrate our proprietary data into the workflow prior to introducing GenAI into production… | Avahi, has the experience to help you understand how to accomplish this using AWS best practice frameworks and services… | Ignite Jumpstart typically 8-12 weeks |
We would like to see how efficient GenAI could be for our developers… | Avahi, has develop quickstarts and best practice KPIs to aid you with your decision… | Ignition CodeWhisperer PoC 30-days |
We’ve conducted successful PoCs with our own data and are ready to move into production, but we’re having trouble understanding how to deploy at-scale and manage cost… | Avahi has worked with customers to design and cost-optimize your model training, inference and the ongoing MLOps environment… | Custom SOW for at-scale, production-grade solution on AWS TBD |
- Workshops – traditionally ½ day online events. Customers can schedule 1on1 onsite workshops if applicable
- QuickStart PoCs
- AWS Marketplace
- AWS Partner Portal ACE
- Traditional marketing channels: Email | Social Media | SEO