--- title: "Will AI in health mean legible prescriptions? " newsletter: "MLOps Community" date: 2024-07-25 source: https://aaif.live/newsletters/mlopscommunity/2024-07-25-will-ai-in-health-mean-legible-prescriptions --- # Will AI in health mean legible prescriptions? *Plus, piledriving AI hype, be better than Bixby, Whisper it...part 2 of semantic search blog, hack-a-team, and hidden gems.* *MLOps Community — Agentic AI Foundation, 2024-07-25* ## AI in Healthcare // Eric Landry // MLOps Podcast #249 Is developing RAGs giving you cold sweats and heart palpitations? In this episode, Eric shares his experience developing RAG chatbots for healthcare. He talked about the challenges in early AI work with limited compute resources, building cloud-agnostic AI platforms, and optimizing model deployments. His current work involves engaging underserved communities and using curated knowledge bases to ensure high-quality data. We also chatted about the need for ethical considerations, real-time monitoring, and strategies to limit hallucinations and improve healthcare outcomes. This podcast could literally be a life-saver. Video [https://home.mlops.community/home/videos/ai-in-healthcare] || Spotify [https://open.spotify.com/episode/1eptohM0SC10I56mnX5Aj7?si=5a7ef010e3cd4328] || Apple [https://podcasts.apple.com/us/podcast/ai-in-healthcare-eric-landry-249/id1505372978?i=1000662764085] [https://podcasts.apple.com/us/podcast/ai-in-healthcare-eric-landry-249/id1505372978?i=1000662764085](https://podcasts.apple.com/us/podcast/ai-in-healthcare-eric-landry-249/id1505372978?i=1000662764085) [https://home.mlops.community/home/videos/ai-in-healthcare](https://home.mlops.community/home/videos/ai-in-healthcare) [https://open.spotify.com/episode/1eptohM0SC10I56mnX5Aj7?si=5a7ef010e3cd4328](https://open.spotify.com/episode/1eptohM0SC10I56mnX5Aj7?si=5a7ef010e3cd4328) ## AI Operations Without Fundamental Engineering Discipline // Nikhil Suresh // MLOps Podcast #250 Struggling to wrestle the AI hype and realistic use cases? Nick piledrives the message home, sharing his frustrations about the implementation of AI without clear objectives or proper groundwork, and how it leads to unproductive roles and unrealistic expectations. We also chat about his transition from data science to data engineering due to inadequate infrastructure in many businesses, and the impact of GPT-3. A bodyslam of an episode! Video [https://home.mlops.community/home/videos/ai-operations-without-fundamental-engineering-discipline] || Spotify [https://open.spotify.com/episode/5LQ8GZYfcK2sp3AfJiDuNP?si=7adddb3ab2f34c45] || Apple [https://podcasts.apple.com/us/podcast/ai-operations-without-fundamental-engineering-discipline/id1505372978?i=1000663114747] [https://podcasts.apple.com/us/podcast/ai-operations-without-fundamental-engineering-discipline/id1505372978?i=1000663114747](https://podcasts.apple.com/us/podcast/ai-operations-without-fundamental-engineering-discipline/id1505372978?i=1000663114747) [https://home.mlops.community/home/videos/ai-operations-without-fundamental-engineering-discipline](https://home.mlops.community/home/videos/ai-operations-without-fundamental-engineering-discipline) [https://open.spotify.com/episode/5LQ8GZYfcK2sp3AfJiDuNP?si=7adddb3ab2f34c45](https://open.spotify.com/episode/5LQ8GZYfcK2sp3AfJiDuNP?si=7adddb3ab2f34c45) ## Job of the Week [https://jobs.ashbyhq.com/Deepgram/5b193c65-5cac-43e8-aaa9-a10882b5f7c4?utm_source=NyMoPd70YW](https://jobs.ashbyhq.com/Deepgram/5b193c65-5cac-43e8-aaa9-a10882b5f7c4?utm_source=NyMoPd70YW) ## Databricks Assistant Through RAG // Neha Sharma // MLOps Community IRL Meetup #87 San Francisco ## MLOps Community IRL Meetup From Clippy to Bixby, building assistants can be tough. This talk covers the development and usage of the Databricks Assistant, focusing on key features such as contextual conversation, code authoring, error diagnosis, and SQL query generation. It delves into evaluation strategies, including helpfulness benchmarks, side-by-side user preference tests, and tracking accept rates and AI-generated code engagement. Looking to the future, it discusses fine-tuning models to improve quality and creating personalized user experiences. 📎It looks like you're going to watch a video. Enjoy! Watch it here [https://home.mlops.community/home/videos/databricks-assistant-through-rag] [Watch it here](https://home.mlops.community/home/videos/databricks-assistant-through-rag) Semantic Search to Glean Valuable Insights from Podcast Series Part 2 [https://home.mlops.community/home/blogs/semantic-search-to-glean-valuable-insights-from-podcast-series-part-2] Add this to The Empire Strikes Back, T2: Judgement Day, and The Godfather Part II as a sequel that delivers. Continuing from part 1 [https://home.mlops.community/home/blogs/semantic-search-to-glean-valuable-insights-from-podcasts-part-1], this blog details generating transcripts using OpenAI's Whisper model, setting up the database, and ingesting the text files. Cohere's embed-english-v3.0 model was chosen for its accuracy and multilingual support and the transcripts were chunked before embedding. The blog explains how embeddings were saved in a numpy file and ingested into ApertureDB using a descriptor set and Parallel Loader for efficient batch processing. Give it a read—it's even better than Breakin' 2: Electric Boogaloo. With thanks to Sonam Gupta for their contribution. How I built KinConnect to win 2000$ of Fireworks.ai! [https://home.mlops.community/home/blogs/how-i-built-kinconnect-to-win-2000dollar-of-fireworksai] Sometimes, the toughest part of a hackathon is finding the right teammates. This blog details the creation of a profile matching system using Google Forms for data collection and Pipedream for event-driven backend integration. Profiles were matched using a combination of Fireworks AI models and MongoDB's hybrid search capabilities. The project underscored the importance of synthetic data generation for testing and refining, and prompt experimentation with tools like promptfoo. Read it now, who knows when your next ‘hidden gem’ of a hackathon will be! With thanks to Nehil Jain for their contribution. [Semantic Search to Glean Valuable Insights from Podcast Series Part 2](https://home.mlops.community/home/blogs/semantic-search-to-glean-valuable-insights-from-podcast-series-part-2) ## Hidden Gems ## IRL Meetups LA [https://lu.ma/z1akzuur] - July 25 Denver [https://mlops.community/event/pizza-pints-expert-speaker-intro-to-machine-learning-toolkits-with-kubeflow/] - July 30 Bucharest [https://lu.ma/gbi3flph] - July 31 San Jose [https://lu.ma/sjzmq522] - August 6 (📣shout out to Google Cloud) [LA](https://lu.ma/z1akzuur) Interested in partnering with us? Get in touch: partners@mlops.community Thanks for reading. See you in Slack [https://go.mlops.community/slack], YouTube [https://www.youtube.com/channel/UCG6qpjVnBTTT8wLGBygANOQ?view_as=subscriber], and podcast [https://home.mlops.community/public/content/] land. Oh yeah, and we are also on X [https://twitter.com/mlopscommunity]. The MLOps Community newsletter is edited by Jessica Rudd [https://www.linkedin.com/in/jmrudd/]. --- Source: https://aaif.live/newsletters/mlopscommunity/2024-07-25-will-ai-in-health-mean-legible-prescriptions