Local large language models are having a moment. Much as I love online AI models like Perplexity, I care about my data and have been using local AI models to boost productivity. Over the past few months, I've been running these models entirely on my own machine, and not a massively powerful one at that, and it has completely transformed how I work. Using a local model via Ollama instead of relying on cloud services has many benefits. Chief among those, of course, is offline access and privacy. While there are many ways to amplify your productivity using AI, here are three ways I'm using a local LLM to boost mine.

Knowledge management

Your personal research assistant

This one's a no-brainer, but one of my favorite use cases is as an accompaniment to my personal knowledge management app. While I'm big on Obsidian and Notion, a local LLM can go a long way towards helping you manage that information better. For example, I make a lot of meeting notes, articles, research documents, and, usually, I'd have to go through these manually. One document at a time. Now, I can feed all the documents into a tool like TL/DW or Obsidian Copilot and get it to process them using a large local language model. This lets me create summaries, pull out action points, and even have a conversation with my own documents for better learning. It's basically a cheat sheet for creating Cliff Notes, and I love it. Of course, there are the usual benefits, like retaining ownership of all the data, not just my documents, but also all the summarization data created by the LLM. This is data that I cannot afford to have on the internet as training data, so public LLMs are out of the question. Having my own LLM also lets me play around with formats, like creating flashcards for learning or getting them to create a summarized digest of multiple research papers that I can read up on using my Kindle. It's the best personal research assistant I've ever had.

Content management

Saving time without losing control and context

The second workflow is perhaps the most controversial. As a journalist, I'm vehemently opposed to the idea of using AI slop. But I do feel that AI has a place in modern workflows to accelerate productivity. Drafting a boilerplate contract email? Get the local LLM to create a template for you. Need to create an outline for talking points to cover in a pitch deck? LLM to the rescue. Or how about asking it to poke holes in a write-up? Grammar checks. Ideating headlines. Making sure the best SEO practices have been followed. These are all tasks that an LLM excels at. It doesn't need creativity; it just needs the model to hit check marks. As in the case of knowledge management, this is original data that I do not want on the internet, so a local LLM works best. Yes, a local LLM is nowhere near as good at text generation, but I don't need it to be. On average, I save at least ten to fifteen minutes in research and basic outlining because I can get the LLM to tell me if I've hit all the relevant talking points in my first draft.

Home automation

When Home Assistant gets an AI brain

I love Home Assistant. It's the center of my smart home. But as powerful as Home Assistant is, it truly comes into its own when you pair it with an LLM. With recent updates, Home Assistant has made the task incredibly easy, which means I can get it to generate everything from routines to automations. When in a pinch, I've asked the local LLM to create a boilerplate YAML file that I can further work with for my dashboard. This is data I absolutely do not want exposed to the internet, so the ability to connect with an Ollama instance on my computer offers me both features and privacy. It's a win-win situation in my books.

Adding a local LLM is the ultimate productivity hack

Of course, there are dozens of other ways to use local LLMs to your advantage, but for me, these three use-cases have stood out the most to accelerate my productivity. It's transformed how I handle everything from my personal life to my home and my work. I'm now able to gather information faster than I ever could, and wrangle that into a form that is best suited for the task at hand. The local LLM stands by me as I write and pokes holes in my work to make it better, while also offering basic grammar checking to make my editor's life easier. And to make my life easier, it taps into my smart home to offer more contextual advice. All of which to say, adding a local LLM to my daily setup has been the best productivity boost I could've hoped for. Not only am I faster than ever, but it also gives me more time and mental space to focus on everything else beyond work.