Sarah Floris, MS's Linkedin Analytics

In 2023, I was writing a lot on LinkedIn and Substack. Then, suddenly, I stopped. No explanation, no heads-up, nothing. Looking back, that was not fair to you—but I also needed to get away. So here is an honest update on where I have been and what has been going on. 1. My mental health is a work in progress (but mostly sucks lol) I did not have much of a childhood growing up, and that is something I carry with me and work through continually. On top of that, I put my husband through school, dealt with the trauma of being laid off several times, and endured toxic work environments (not at EquipmentShare! Let me know if you want a referral!). All of that had taken its toll. 2. Transitioning into machine learning
 At first, I was excited, but soon imposter syndrome hit hard—and it still lingers at times. I have been pushing through, learning as I go, and my data engineering background has been a huge asset. 3. The noise around AI
 So many people talk about AI without actually working in the field. Oversimplified takes, exaggerated claims, and endless information made it exhausting to find what I am looking for. The disconnect from reality made it hard to stay engaged. 4. My health
 Being diagnosed with chronic illnesses was probably a pivotal moment for me. The tiredness and pain pushed me to recognize my own limitations in ways I hadn't anticipated. It's isolating, and admitting the challenges I have faced has not been easy. I'm not sharing this to exaggerate my journey or wrap everything up perfectly. The past few years have been challenging, and they continue to be. Some days are more manageable than others, yet I'm still navigating through it all. I feel a responsibility to you—the followers who've supported me—to be transparent about my whereabouts. Thank you for sticking with me, even when I went quiet. Your support means more than I can put into words. ----------------------------------------------- Moving forward, I want to—and will—be more present. Here is why. Despite having a strong network of support, I often struggle to articulate my emotions or track my progress. I have been searching for a tool that would assist me in processing my thoughts and emotions, fostering personal growth, and evolving into a better version of myself—all while ensuring my privacy and security. Most AI tools available do not prioritize these aspects, so I decided to create one myself. That is why I am building Thunderbuddies.ai, a platform focused on providing a safe space to grow and mature. I will also be sharing my journey of building this tool through my Substack, Bootstrapping AI, alongside what other folks are building. Check it out 👇 #data #ml #ai #datascience

AI tools can be extremely useful but be wary of what tools you are using and how they use your data. Most of the time the data is available whenever they want for however long they want. ChatGPT does have an option for Enterprise customers to have the chats available for up to 90 days, but this is not set by default. #data #ml #datascience

Any Snowflake folks in my network who know about snowpark vs regular SQL? No, I do not want to start a war. Lol

I know we have heard of LangChain and Llama-index, but have you heard about LocalAI? I am looking at tools that prioritize privacy first and LocalAI definitely meets that requirement. So what is LocalAI? 👉 LocalAI is a self-hosted solution that lets you run large language models, audio, and image generation tools entirely on your own infrastructure. 👉 Designed as a drop-in replacement for OpenAI, LocalAI supports the same REST API spec, making it easy to integrate with existing applications built on OpenAI's ecosystem. Key features: 1. OpenAI-compatible API for easy integration 2. Supports a wide range of models (LLaMA, Mistral, Whisper, etc.) Runs on CPU or GPU — including Raspberry Pi and air-gapped machines 3. No external dependencies or cloud access required 4. Optimized model loading and caching for fast inference Check it out at localai.io. #opensource #LLM #privacyfirst #ml #datascience