Patricia Diaz-Hymes

@UMich. Isleña | Dominicana. IT Product #Marketing. [All tweets are my own]

[Andrej Karpathy @karpathy]

LLM Knowledge Bases

Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So:

Data ingest:
I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them.

IDE:
I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides).

Q&A:
Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale.

Output:
Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base.

Linting:
I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into.

Extra tools:
I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries.

Further explorations:
As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows.

TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

[Patricia Diaz-Hymes @hashtagpdiaz]

@ShortiiRock @amazon Same

[Patricia Diaz-Hymes @hashtagpdiaz]

@sjoyklein @OlaySkin @JLo @shakira Amen

[Patricia Diaz-Hymes @hashtagpdiaz]

Still not over the #superbowl halftime #latinapower

[Patricia Diaz-Hymes @hashtagpdiaz]

#latinapower #latina #SuperBowl best half time show ever. So proud of being Latina now ❤️ and always.

[Patricia Diaz-Hymes @hashtagpdiaz]

Thank you @CIOsynergy for having @lakesidesoft & me on stage today! #ciohouston

[Patricia Diaz-Hymes @hashtagpdiaz]

Check out our key takeaways from Gartner's Market Guide for Digital Experience Monitoring https://t.co/OQQwH08D6M

[Patricia Diaz-Hymes @hashtagpdiaz]

#digitalexperiencemonitoring #DX #APM #employeeexperience calculate the ROI on monitoring the digital experience of employees!

[Lakeside Software @lakesidesoft]

What can SysTrack do for you? With our free ROI calculator, you can measure the impact our digital experience monitoring tool can have in your organization. Find out what you could be saving today - https://t.co/vXyoxhZRGD

[Lakeside Software @lakesidesoft]

THIS JUST IN: We have been named a finalist for the Digital Workspace category at VMworld! Come learn about our award-winning technology: SysTrack 9.0

[Patricia Diaz-Hymes @hashtagpdiaz]

@DrPizza Yes - so is the free SysTrack Windows Virtual Desktop Assessment. May be of interest to you as WVD moves to GA! https://t.co/QjlvwWafpN

[Lakeside Software @lakesidesoft]

Did you hear? Windows Virtual Desktop is in public preview but is it right for your organization? Check out our FREE tool to help with that... https://t.co/az1al5Swgg

#WVD #Citrix #Azure #WindowsVirtualDesktop #Windows10

[Patricia Diaz-Hymes @hashtagpdiaz]

Is WVD right for you? Check out our latest post on the topic! https://t.co/QjlvwWafpN

[Patricia Diaz-Hymes @hashtagpdiaz]

🙌🏽

[Thomas Poppelgaard @_POPPELGAARD]

Great to see mix gender diversity at #IGELDisrupt #EUC #womanintech

[Patricia Diaz-Hymes @hashtagpdiaz]

#behumble

[Gilda Giannina Doria @gilda_doria]

"If you aren’t grateful for what you already have, what makes you think you would be happy with more." #tuesdaythoughts thanks @ashozz for the share!

[CHEN PR @CHEN_PR]

Looking ahead to 2019, @lakesidesoft's @Heather_Bick shares 5 predictions on how #IT & #technology may shape the new meaning of work -- from an increase in remote offices to growing IT pressure, she tells all via @VMblog: https://t.co/EiDlvDfOW5

[Lakeside Software @lakesidesoft]

We all know artificial intelligence (AI) is a hot topic — but beyond the buzzword, have you ever wondered how IT departments are actually adopting AI technologies to improve on their operations? Check out the article by @hashtagpdiaz on@APMdigest https://t.co/dcZ1bMpbod

[Lakeside Software @lakesidesoft]

We're excited to announce SysTrack is available to customers of @VMware Cloud™ on #AWS. Get the full scoop in our latest blog. https://t.co/FqjS643fGz #EUC #vmware #cloud

[Lakeside Software @lakesidesoft]

Over 98% of large enterprises with 1000+ employees say that on average, a single hour of downtime per year costs their company over $100K --- Good news: You can prevent it and be proactive with a self-healing help desk. https://t.co/2Rxp4fBpdI #ITSM #TuesdayMotivation

[Lakeside Software @lakesidesoft]

We're at #CIODetroit! Where our very own, @hashtagpdiaz, is talking about how to mend the rift between end-users and IT! Be sure to stop by our booth after our session for an awesome prize and get an inside look at our newest #SysTrack release.

[Tal M. Klein @VirtualTal]

In case you missed this from me yesterday. I wrote a little thing on how Level 0 support (self service/self heal) is the thing behind the “shift left” movement.