Matt Kalan

Expert in solving business problems with technology especially for enterprises - currently solution architect at MongoDB guiding customers in data management

[Matt Kalan @MatthewKalan]

Useful to have one term that includes all levels of autonomy in #chatbots/#agents/etc #AIML #appdev #genai

[LangChain @LangChain]

🧠What is a "cognitive architecture"?

This is a term we've been using a bunch recently. All credit for coming up with it goes to @Altimor

We tried to write down a bit what exactly we mean by this & how to think about and explore different options here

https://t.co/az589jwCDO

[Matt Kalan @MatthewKalan]

@langchain @Altimor I like that this is one term that includes levels of autonomy in #chatbots/#agents/etc #AIML #appdev #genai

[LangChain @LangChain]

🧠What is a "cognitive architecture"?

This is a term we've been using a bunch recently. All credit for coming up with it goes to @Altimor

We tried to write down a bit what exactly we mean by this & how to think about and explore different options here

https://t.co/az589jwCDO

[Matt Kalan @MatthewKalan]

#AIagents can use #LLMs to make smart decisions and take action. Check out this tutorial for building a research assistance agent w/ @MongoDB, @FireworksAI_HQ, & @langchain https://t.co/m4ehPgjm8A #appdev #genai #aiml

[Matt Kalan @MatthewKalan]

@MarcusForPeace Was this related to something specific he said? 😄

[Matt Kalan @MatthewKalan]

Benny Cheng, Co-founder of @FireworksAI_HQ, discussing their value for #inference and #finetuning in the #GenAI app ecosystem - from @theCUBE at #mongodblocal in NYC last week on
https://t.co/o8h0xlYMCN #aiml #appdev

[Matt Kalan @MatthewKalan]

@TheDonester @pawel_mano I so want that lego! I guess I have to connect image generation to a 3D printer and I could!

[Matt Kalan @MatthewKalan]

@TheDonester Welcome to the club!!!

[Matt Kalan @MatthewKalan]

Really nice to see @Retool in its survey on the state of #AI to find that the @MongoDB Atlas cloud service has the best NPS scores for its #vectorsearch features and is the 2nd most deployed #vectordatabase
https://t.co/odgIYI7wWp #GenAI #LLM #AIML

[Matt Kalan @MatthewKalan]

@langchain just released Templates and @MongoDB posted a #RAG template using Atlas Vector Search and @OpenAI for building a #chatbot using your own private data.

Read about it in the MDB blog post:
https://t.co/HOkdaV3RGG #LLM #AIML #retrievalaugmentedgeneration #appdev #genai

[Databricks @databricks]

We've added a ton of cutting-edge advancements in #UnityCatalog: #Lakehouse Federation, Governance for AI (including Model Registry), AI-powered Governance (Lakehouse Monitoring, Lakehouse Observability) & more!

Get a full breakdown here ⬇️
https://t.co/Nmupr8hdbF

[Matt Kalan @MatthewKalan]

@Databricks beefing up its ability to help customers with creating their own #LLMs for #GenAI - nice combination of #opensource driven companies #AIML #data https://t.co/HfvzMUznDr

[Matt Kalan @MatthewKalan]

Nice preview of @MongoDB's biggest #mongodblocal event of the year in NYC this Thu Jun 22, to showcase MDB as the foundation for the next generation of apps including for #GenAI
#appdev #databases https://t.co/UlbTxIY2d2

[Matt Kalan @MatthewKalan]

I had the thought of whether #LLMs could give a jumpstart in building #personalization without training a model, and apparently there are many posts on #zeroshot #recommendationsystems. This seems like the future, using prebuilt LL…https://t.co/gi9y3H9nKg https://t.co/zW89Jsucc3

[Matt Kalan @MatthewKalan]

@EdenEmarco177 @langchain @hwchase17 Could approaching it this way for arbitrarily large quantities of text be reasonably accurate? The mapper would return the nearest vector for each block of text and the reducer would find the midpoint of all the vectors

[Matt Kalan @MatthewKalan]

@temporalio code is cut off on your home page for me

[Matt Kalan @MatthewKalan]

Nice post from the @MongoDB product manager for #AIML Prakul A. on enabling #chatgpt on private data with @LlamaIndex #LLM #GenAI https://t.co/VKIe2LDlq5

[Matt Kalan @MatthewKalan]

Good reminder from @TechCrunch to stay focused on the many near-term risks of #ai plus the #extinction ones

"This drumbeat of hysterical headlines has arguably distracted attention from deeper scrutiny of existing harms. Such as…https://t.co/QrZdAnmJqd https://t.co/F8YSBms2NI

[Matt Kalan @MatthewKalan]

Certainly novel to use #LLMs to explain other LLMs - at least good to see focus on explaining the models as that will be necessary to justify otherwise black box decisions #AIML might make
https://t.co/KYhmBbbXzJ #openai https://t.co/nYTvmCvQKZ

[Matt Kalan @MatthewKalan]

@mjasay You're just trying to fail fast I guess!

[Matt Kalan @MatthewKalan]

My humble colleague @mjasay @mjasay announcing he might have been wrong and the #AI singularity is [80%] here. Human + AI gives massive productivity for #appdev when you look at benefits across the SDLC, for writing first draft cod…https://t.co/wW2P6A4FoO https://t.co/WC4I4KMLg6