Build and orchestrate serverless generative AI applications | Mohammed Fazalullah | Conf42 CN 2024

Read the abstract ➤ https://www.conf42.com/Cloud_Native_2024_Mohammed_Fazalullah_build_orchestrate_serverless Other sessions at this event ➤ https://www.conf42.com/cloud2024 Support our mission ➤ https://www.conf42.com/support Join Discord ➤ https://discord.gg/DnyHgrC7jC Chapters 0:00 intro 0:24 preamble 0:35 genai is the answer. what was the question? 1:31 accelerate drug discovery 1:41 tailor investment strategies 1:52 enhance medical images 2:03 create product listings 2:30 genai stack 4:49 amazon bedrock 6:44 how do i access foundation models? 7:18 integrating with amazon bedrock 8:14 sequencing is hard 11:02 aws step functions 12:55 directly compose apps from over 220 aws services and 10k api actions 14:05 control service integrations with call patterns 15:15 why direct integration is powerful 17:48 machine learning and genai 18:46 building a genai application 19:11 optimized integration for amazon bedrock 20:14 demo 27:02 requirement 1: generate title and description 27:18 amazon states language (asl) for amazon bedrock 28:22 building a genai application 28:37 how do i access foundation models outside aws? 28:49 accessing a public api 29:46 public https api integration 31:26 asl for https endpoint 32:19 requirement 1: generate multiple titles 32:57 what if something goes wrong... 33:57 how do you handle hard failures? 34:38 redrive 35:22 execution event history for easy observability 36:01 building a genai application 36:43 wait for callback - human / external process 37:26 requirement 2: human feedback loop 37:46 building a genai application 37:52 requirement 3: generate avatar 38:35 prompt chaining 39:14 architecture 40:09 demo 42:56 resources 43:18 thank you