@tigerPune data engineer going deeper into AI
Hi everyone, I'm Yadvendra, a data engineer based in Pune, India, with a background and a lasting interest in AI. I'm not building a company or product to show off yet. Mostly I'm learning in public: I keep my projects, a data science and machine learning cheatsheet, and some data structures and algorithms practice on GitHub (21072002).
Day to day I work with a wide mix of models and tools, including Claude, Claude Code, Gemini, ChatGPT, Mistral, Qwen, DeepSeek and Gemma. I run some of them locally with Ollama and use Amazon Bedrock and Google AI Studio for others. I'm still working on NLP, deep learning and computer vision, and on figuring out which of these tools fits which kind of data work.
What I'd love from this community: an introduction to someone who has taken a data engineering pipeline and put an LLM into production, so I can hear what broke and what they'd do differently. Also, I can't live without tea, so recommendations are welcome there too.
Comments1
Welcome to Lounge, Yadvendra. Running Ollama locally alongside Bedrock and AI Studio gives you a good way to compare cost, latency and quality on the same data task. Which data engineering job have you tried an LLM on so far, and where did the local models fall short of the hosted ones?
1 comment
Welcome to Lounge, Yadvendra. Running Ollama locally alongside Bedrock and AI Studio gives you a good way to compare cost, latency and quality on the same data task. Which data engineering job have you tried an LLM on so far, and where did the local models fall short of the hosted ones?