Harrison Chase
@hwchase17
Followers
73K
Following
12K
Media
513
Statuses
13K
@LangChainAI, previously @robusthq @kensho MLOps ∪ Generative AI ∪ sports analytics
Joined July 2014
OpenAI recently released a guide on building agents which contains some misguided takes. There's a lot of FUD, confusion, hype, and noise around agents. I wrote a blog on how to think about agent frameworks. Includes:. Background Info.- What is an agent?.- What is hard about
90
117
877
🌟privateGPT🌟 - this is sick!!!. I've always had people asking me if it was feasible to use @LangChainAI with open source models, and my answer was "at the moment, not really. ". but @ivanmartit did it!!!!! this is a huge step forward 👏👏👏.
59
402
2K
🤖Generative Agents🤖. Last week, Park et all released “Generative Agents”, a paper simulating interactions between tens of agents. We gave it a close read, and implemented one of the novel components it introduced: a long-term, reflection-based memory system. 🧵
51
343
2K
Baby-AGI by @yoheinakajima is taking the world by storm. Here's an implementation within the @LangChainAI framework, allowing you to easily substitute in other vectorstores and other LLMs. Docs:
52
294
2K
Really excited to publicly announce that:. 1⃣ @ankush_gola11 and I have started a company around @LangChainAI .2⃣We raised a seed round from @benchmark. We wrote a bit about the journey so far, and where we want to take it:.
177
156
2K
LangChain 🤝 AIPlugins. A first open source attempt at using AIPlugins (the same ones ChatGPT is using). s/o @vaibhavk97 for this. Excited to see what other techniques the @LangChainAI community comes up with - it's only the beginning. Docs (Python and JS) in 🧵
44
227
1K
OpenAI recently open-sourced `chatgpt-retrieval-plugin`, a module for creating a retriever based on your own data. Can you combine this with @LangChainAI's 40+ DocumentLoaders?. Yes. Yes you can. Gist for doing so:
27
159
1K
ChatGPT @OpenAI API was released today, and the first of many integrations is now available in @LangChainAI . Python docs (`pip install langchain==0.0.98`): JS/TS docs (`yarn add langchain@0.0.18`):
26
155
1K
Yesterday @karpathy tweeted about using SVMs instead of KNN for retrieval (pros: better results, flexibility; cons: takes longer). Today @RLanceMartin implemented it in @LangChainAI 🚀🚀. Play around with it here!.
26
130
964
⭐️Claude + AI Plugins⭐️. AI Plugins (the ones ChatGPT is using) are usable by ANY language model. It takes a tiny bit of prompt engineering, but here is @AnthropicAI's Claude using them. Code:
25
139
951
Retrieval for QA systems is hard. Vector search is good for capturing semantically similar texts, but often queries specify desired attributes like time, authorship, or other "metadata" fields, which vector search is not great at. Enter. ⭐️SelfQueryRetriever⭐️
40
139
940
🚨Emergency OpenAI Functions Release🚨. `pip install langchain==0.0.199`. ✅ Support for functions in chat model wrapper.✅ Convert @LangChainAI tools to functions. Docs: Next up. I think its time for a new type of agent 😁
56
167
882
My TED talk is out!. I talk about what it takes to build "context-aware reasoning applications", including:. 4 different ways of providing context to LLMs.6 different types of cognitive architectures.
34
124
837
🤖Autonomous Agents & Agent Simulations🤖. Four agent-related projects (AutoGPT, BabyAGI, CAMEL, and Generative Agents) have exploded recently. We wrote a blog on they differ from previous @LangChainAI agents and how we've incorporated some key ideas.
30
165
815
ChatGPT 🤝 WolframAlpha. Give #ChatGPT a Wolfram|Alpha neural implant (as @stephen_wolfram put it) with @LangChainAI. 👏👏 @nickscamara_ for adding this. Collab Notebook to try it out: Screenshot Below
17
160
768
🦜🔗Code Interpreter API. ChatGPT's code interpreter is the hottest thing in the streets. A new project by @Shroominic takes a stab at recreating that functionality locally using OpenAI's apis. Uses CodeBox - a sandboxed env with simple fileIO functionality. Blog & GitHub 👇
23
149
772
The new @OpenAI functions are good for other things besides agents. Another killer use case is extracting structured information from unstructured docs. We've adding support for extraction AND tagging in @LangChainAI - thanks to @fpingham for code and @jxnlco for review. 🧵.
27
107
719
📰GPT-Newspaper. This project combines six agents to autonomously build a newspaper - content, outline, everything! . By the same team (@tavilyai) that created GPT-Researcher. Code:
13
122
687
🐪CAMEL🐪. Communicative Agents for “Mind” Exploration of LLM Society. This paper shows how to put 2 agents in a sandbox with each other and watch them interact. Now implemented in LangChain! (s/o @guohao_li). Original Paper: Docs:
29
135
675
LLMs are pretty good at writing SQL, but still struggle with some things (like joins). 🤯 But what if you use an *agent* to interact with SQL DBs?. In the example below, it tries a join on a column that doesn't exist, but then can see the error and fixes it in the next query
19
68
675
🦜🤖LangManus. You had to know it was coming!. This community effort attempts to replicate Manus using the LangStack (LangChain + LangGraph). Still early innings - but check it out here!.
24
128
690
LLMs are great at understanding text. This allows them to extract structured information from text: to use in forms, for query generation, knowledge base construction, etc. `pip install kor` by @veryboldbagel is the best attempt at this I've seen.
23
93
668
Excited to launch the world premiere of Besides being aesthetically beautiful, it also has some much requested documentation:. ⚙️ Integrations (Python vs JS/TS): ⛳️ Features (Python vs JS/TS):
57
49
674
Want to give your agent access to 20k+ tools?. 🔥@LangChainAI x @zapier🔥. Integration now out in Python and JS. Blog Post: Python Docs: JS Docs:
14
120
654
⭐️Contextual Compression⭐️. We introduce multiple new methods in @LangChainAI to compress retrieved documents w.r.t. the query before passing to an LLM for generation. Inspired by @willpienaar at the "LLMs in production" conference. Blog: 🧵More details:.
20
125
642
How to best build an agent that has access to ALL the ChatGPT Plugins?. IMO, by combining two techniques we recently introduced in @LangChainAI . 🔧 Tool Retrieval.🗣️ Natural Language APIs. Explanation and example notebook in 🧵
9
103
638
🦺RAGxplorer. There's a lot of nuance in RAG, a lot of different parameters to control. It can be really helpful to just play around with parameters and visualize results. @gabchuayz built an AWESOME open source @streamlit app for exactly this.
18
101
564
Agents are here, you just have to know how to build them. Excited to share this VERY detailed blogpost from @unifygtm on how they built their recent account qualification agent. This contains all sorts of nuggets like:.3⃣The three different cognitive architectures they considered
6
78
617
✂️Text Splitting Playground. Chunking text into appropriate splits is seemingly trivial yet very nuanced. Open sourcing a playground to help explore different text splitting strategies. GitHub: Hosted Playground:
16
121
596
I hadn't seen this example until someone opened a PR to put it in the @LangChainAI gallery. but it's pretty sick. DocsGPT: answer questions about the documentation of any project.
11
93
585
📄Documents x LLMs📄. Combining documents with LLMs is a key part of retrieval and chaining. We've improved our @LangChainAI reference documentation across the 5 major CombineDocumentsChains and helper functions to help with clarity and understanding of how these work. 🧵
20
89
567
🚨pip install --pre langchain==0.0.99rc0🚨. Play around with 4 chains built to specifically interact with the ChatGPT API in the proper way (rather than forcing it into the standard LLM wrapper). - ConversationChain.- QA Chain.- VectorDBQA Chain.- ChatVectorDB Chain.
13
48
559
New Flan-UL2 model on @huggingface. time to try it out in @LangChainAI!. Put together a notebook testing it with a standard chain-of-thought prompt, as well as question-answering over specific documents. h/t @ClementDelangue for the idea!.
8
108
526
🚀We're launching "long-term memory" support in LangGraph. At its core, long-term memory is "just" a persistent document store that lets you *put*, *get*, and *search* for memories you've saved. Why so simple?. 🧵
16
70
530
LangChain-AGI 🤖. A (somewhat) facetious attempt to create AGI using the best agent & tools in @LangChainAI. Can currently use search, do complex math, lookup weather, lookup current news, and more. Notebook here:
10
90
522
Gonna beef up the tutorials for how to create your own Chat-GPT over specific documents with @LangChainAI . What types of documents/knowledge bases would people want to have examples for? Eg Notion, Obsidian, webpages, etc.
73
22
514
For the past six months, I haven't checked email directly, but rather have relied on an AI email assistant to triage and draft emails for me. This is an example of an "ambient agent". We're open sourcing the code today, as well as a hosted version for people to try out. 👇
33
67
497
im bullish on long-running, stateful agents. who's building one of those?.
134
19
515
⛓️Chain of Verification. A great new paper from Meta on a prompting technique to reduce hallucinations. 🦜🔗Sourajit Roy Chowdhury implemented this in @LangChainAI **along with some improvements**. 📃And he wrote a blog on it. 🧵Lets dive in (this is why I love the LC community!)
8
112
495
New summarization technique:. ⛓️Chain of Density🍢. Produces 5 summaries iteratively, each one denser and more informative than the prior. See @vimota's excellent thread on it below. Try it out on the prompt hub!.
The latest paper from @salesforce AI, et al. discovers a new prompt called Chain of Density (CoD) that produces more dense and human-preferable summaries compared to vanilla GPT-4. I gave it a try for a few articles and it's really solid. Pasted it here if you want to try it 👇
12
70
475
How to speed up "chat-your-data" applications while retaining final answer accuracy?. 🫙Use a cheaper/faster model (gpt-3.5) to create the condensed question.💬Use a better but more expensive model (gpt-4) for final response. Thanks to @cristobal_dev for highlighting!. 🧵
15
59
473
✒️OpenAI's Bet on a Cognitive Architecture. I wrote a blog on how GPTs & Assistant API represent a bet OpenAI is making on building a specific cognitive architecture. As well as some thoughts on what cognitive architectures currently work and why. Blog:
15
93
463
Integration 2/n: GPT-4. If you are lucky enough to have access to GPT-4, using it in @LangChainAI is quite simple: just update the model name. ```.from import ChatOpenAI.llm=ChatOpenAI(model_name="gpt-4").```
10
57
448
🎉New @DeepLearningAI_ class🎉. I had so much fun teaching the last one with @AndrewYNg I had to return for a follow up. This one is a deep dive on the most popular applications of LLMs to date: using them to chat with your data. What do we cover? 👇.
24
96
455
If you're looking for a project to distract you on this weekend. may I suggest building a research assistant. Recently filmed a long-form YouTube tutorial on building one from scratch. Covers LCEL, LangSmith, parallelization, retrievers.
56
74
451
A similar "Plan-Execute" agent framework that @yoheinakajima has been using to create Baby-AGI is also useful for interacting with large and complex OpenAPI specs. Here an agent plans and executes 5 different API calls to complete a user request. Docs:
16
77
438
With the release of open source models like StableML by @StabilityAI, Dolly by @databricks and Camel by Writer, we've heard an increase in demand for running these models in @LangChainAI . We recently revamped our docs around @huggingface to highlight how easy this (links 👇)
12
80
447
We just put all the @LangChainAI webinars on Youtube!. The best part is, this now means we can use LangChain (in 12 lines of code) to do question/answering over the LangChain webinars 🤯 How circular. Gist:
20
63
441
🚀OpenAI InstructGPT 3.5 model. OpenAI just quietly released a version of their 3.5 models available via the old completion endpoints. This is nice because chat models are sometimes chatty which make them less ideal for agents, where you want precise steerable outputs. Example 👇
14
72
431
well it happened. ✨the chat-langchain app is completely reproducible in javascript✨. data ingestion, text splitting, embeddings, vectorstore, llms, chains. all through langchainjs. chat-langchain repo: langchainjs repo:
18
48
441
✨Agents + Vectorstores✨: a powerful combo. Can be used to:.🍴 route questions between MULTIPLE indexes.⛓️ do chain-of-thought reasoning with proprietary indexes.🔧 combine proprietary data with tool usage. Here's how to use them together in @LangChainAI 👇.
12
50
442
🤖Agents from scratch. We've rewritten all our 8 agent types using LangChain Expression LangChain and prompts from the Hub. This makes them more modular, understandable, and therefor more customizable. This customizability is crucial for teams looking to go to production. Long 🧵
18
67
442
🤝LangChain x Hugging Face x Gradio🤝. Wouldn't it be cool if you use all the @Gradio apps on @huggingface as tools in a @LangChainAI agent? Thanks to @freddy_alfonso_ you can!. Full Blog: Notebook: Repo:
10
107
439
Introducing LangChain (🦜🔗): a python package aimed at helping build LLM applications through composability. `pip install langchain`. A thread on why we built this, what you can do with it, where it's going, and other tools in the ecosystem:. 🧵.
10
50
443
🚨 Watch how I can run up a $1000 bill with a single call to a poorly protected LLM app 🚨. Prompt injection attack against an agent: tricking it into repeatedly calling the LLM and SerpAPI, quickly racking up costs
11
50
427
Lance (@RLanceMartin) has done a TON to bring advanced retrieval topics to @LangChainAI and make them easily approachable and understandable. This image he put together for our "RAG from Scratch" YouTube series is absolutely 🔥. (YouTube:
11
86
420
Finally merging in proper ChatGPT support to @LangChainAI . Introduced a bunch of new abstractions. For some color on our thought process, please see: To get started, see:. Python Docs: JS/TS Docs: 🧵.
12
60
425
🗺️ Plan-and-Execute Agents 🗺️. Inspired by BabyAGI and the recent Plan-and-Solve paper, we're introducing a new type of @LangChainAI agent. We think these are better for more complex tasks, at the cost of more calls to the LLM. Blog: 🧵.
19
77
419
yeah deep research is great. but have you ever wanted it open source, with swappable models, and able to research over your own data?. GPT-Researcher is exactly that - the leading OSS AI Researcher project.
69
63
425
pretty bad advice here from openai. there are 27 libraries like "Agents SDK" (i would put the original LangChain in this camp!) and none of them are reliable enough to get to production for 99% of use cases. ✒️blog coming this weekend.
46
21
430
🦙LangChain x Llama🦙. Llama integration is finally here! . Not only with the original llama-cpp library, but also with @nomic_ai's gpt4all model (an assistant-style large language model with ~800k GPT-3.5-Turbo Generations based on LLaMa).
Rather large 🦜🔗0.0.131 release!. 🆓GPT4all model (@nomic_ai).🦙Llama-cpp model.⏹️Support for @qdrant_engine local db.🌲Zilliz cloud (@milvusio) Vectorstore support.📧New OutlookMessage Document Loader.🕸️New Selenium Document Loader.🪟 Support for SQL views in SQLChain. 🧵.
8
63
410
It's been one of the more requested topics for a webinar, and it's finally here. 🤖Agents🤖. Featuring:. @yoheinakajima (recently doing a lot of work on baby-AGI).@mbusigin (code agents, cognosis). Next Wednesday, 9am PST:
11
52
411
If MCP is Zapier… won’t the value accrue to the client (not the integrations)?. Who is building the best application to **consume** MCPs?. Or am I thinking about this wrong.
Is MCP just Zapier without UI but powered by a prompt?.
102
18
417
❓What's next for AI agents. I was lucky enough to have the opportunity to talk at the recent @sequoia AI ascent on what was next for AI agents. I talked about three things:. 🗺️Planning.🖥️UX.🧠Memory. Check out the full video here:
13
73
409
yeah you've seen a @LangChainAI agent do math, but have you seen it think about what the best taylor swift song is and automatically add it to your spotify playlist?. IFTTT webhooks now available as tools - in both python AND typescript. s/o @upster for idea and execution
5
41
395
🌲Multi Vector Retriever. The basic idea: you store multiple embedding vectors per document. How do you generate these embeddings?. 👨👦Smaller chunks (this is ParentDocumentRetriever).🌞Summary of document.❓Hypothetical questions.🖐️Manually specified text snippets. Quick 🧵
10
70
382
🔀Router Chains🔀. A simple (yet much requested) abstraction that started with a @ShreyaR pr months ago and is finally in @LangChainAI!. - Router Chain does classification to choose sub chain to use.- Call the selected chain with that input. Lots of potential use cases!. 🧵
20
51
374
What I LOVE about this example is it shows how easy it is to create a CUSTOM retriever class and use it within LangChain. As you go deep on a problem, it's highly likely you'll want some custom retrieval logic to eek out the best performance.
Tutorial: How to build an e-commerce chatbot using #OpenAI, @Redisinc, and @LangChainAI -- useful just if you want to see how this sort of system works.
3
47
371
🕓Ambient Agents. One of the UX patterns I'm most interested in is the idea of "ambient agents" working in the background for us. This can be super powerful. but needs some UX tricks to get right. So I wrote about it!. What are some considerations here?
30
40
373
🧠 Motörhead. Motörhead by @Metal_io is a new memory and information retrieval server for LLMs. Can be used to persist conversational history in @LangChainAI . Motorhead Repo: Python Docs: JS Docs:
3
49
355
🌳Tree-of-Thought. A new reasoning method, originally by @ShunyuYao12, implemented in 🦜🧪langchain_experimental by Vadim Gubergrits. Paper: Docs: LangSmith Trace:
5
60
351
Really excited to share a deployed chatbot grounded to answer questions about LangChain's documentation. ChatBot: Blog: Source Code: HuggingFace Space:
14
42
358
This is my favorite thing we've launched in a while. Been using it for a week and has already changed how I build LLM apps. Build your agent in code, then point LangGraph Studio at it. We'll visualize the agent, let you interact with it, and let you modify the state directly. All
15
47
357
Another way to think about LangGraph is that it is an easy way to create **State Machines**. state machine = labeled, directed graph. Thinking of complex LLM systems as state machines (rather than autonomous agents) can often be useful to enforce more control
15
41
352
⭐️Want an open source version of OpenAI's Operator?. There's a great open source project called Browser Use that does similar things (and more) while being open source. Allows you to plug in any model you want. Love to see open source leading the way🚀.
14
70
357
Our webinar on agents starts in 1 hour. It's the most popular webinar we've hosted yet, so we had to bring in the best possible moderator: @charles_irl . Come join Charles, myself, @ShunyuYao12, @mbusigin and @yoheinakajima for some lively discussion :).
18
47
343
🔧Structured Tools🔧. Agents are all about the tools you give it. Tools in @LangChainAI used to just take a single string input. In our new release, tools can now take multiple inputs. We also introduce a new agent type for these tools. Blog: 🧵.
15
48
344
Many startups focused on serving open source models have exposed APIs that are interoperable with OpenAI. This makes it easy for anyone to try them out by just changing the API base. We added better support for this in @LangChainAI in the most recent release (h/t Yuze Ma) 🧵
17
46
326
✨LangChain x Weights & Biases✨. @weights_biases is the canonical MLOps/experimentation platform. Excited to announce that you can now use it to track your @LangChainAI runs!. Thanks to Anish Shah for working to get this in. Links 👇
4
61
334
GPTs, but with:. 🧠Bring-your-own-model (OpenAI, Azure, Anthropic, Bedrock).🌲Bring-your-own-retrieval (document loading, chunking, embedding, vectorstore, retrieval algorithm).⚒️Bring-your-own-tools.💬Bring-your-own-chat-history. Got a little "research preview" hosted as well.
🦜🤖OpenGPTs. Some big updates to OpenGPTs - an open-source, fully configurable GPTs experience. You can now:. 📁Upload files to a retrieval tool (with full configurability over the ingestion, vectorstore, and retrieval used).🌐Share public bots that you've created.🛠️Use more
20
28
338
coolest client that integrates with MCP that ive seen.
@hwchase17 You should try @tryhighlight - you can just @ any MCP server from anywhere. Runs natively. With context grounding so it can use what you’re looking at as inputs and outputs.
12
39
341
With the ChatGPT release, we've taken the opportunity to rethink memory in @LangChainAI . Main change: allow memory objects to return `List[ChatMessage]` (rather than strings). Also made memory more modular so can be used outside of chains. 🧵 w/ explanation and links.
11
35
330
tell me why your agent framework is better than langgraph. i'm writing a blog (aiming to come out tmrw) about agent frameworks and i want to be as unbiased as possible. but so far i don't have many examples things other agent frameworks do that langgraph doesn't 🤷♀️.
what is something other agentic frameworks do that langgraph doesnt?. not abstractions, but functionality. currently the only one i can think of is dspy by @lateinteraction (which does optimization). what am i missing?.
74
17
342
When connecting LLMs to your data, semantic search is a great first pass - but to get even better performance you can turn to more advanced techniques:. ✒️Contextual Compression.🤳Self Query.🕐Time weighting. Going live in 30m to talk about this and more!.
8
55
329
❓What is an agent?. I get asked this question a lot, so I wrote a little blog on this topic and other things:.- What is an agent?.- What does it mean to be agentic?.- Why is “agentic” a helpful concept?.- Agentic is new. Check it out here:
15
43
336
IMO this is the future for a lot of GenAI apps. Automatically achieve better & more personalized LLM apps by:. - Log all LLM outputs.- Log feedback for those outputs.- Use feedback to build datasets.- Use datasets as few shot examples. Want to do this for your app? Shoot me a DM.
⚙️How Dosu used LangSmith to achieve a 30% accuracy improvement with no prompt engineering. One of the goals of LangSmith is to help teams set up a data flywheel. By capturing LLM outputs alongside user feedback, we can help developers automatically use that feedback to improve
19
39
326
text-to-sql is a really common use case. read here how LinkedIn rolled out a production text-to-sql bot to everyone at the company. the blog linked below is pretty technical, so can definitely learn some things!
🏆Top 5 LangGraph Agents in Production 2024. #3: LinkedIn. While "agents" are the buzzword of the moment, agentic apps built with LangGraph have been in production throughout 2024. Among those who shared insights publicly, we're doing a countdown of our 5 favorite. Next up,
6
40
327
Really excited to announce something new:. ✨LangChainHub✨. A place to share and discover @LangChainAI prompts, chains, and agents. Blog: Hub: Excited to see what gets put on here 😍.
9
47
322
🔥FLARE🔥. Forward-Looking Active REtrieval augmented generation. A new retrieval augmented generation method from Zhengbao Jiang, @luyu_gao et all. Now in @LangChainAI - slightly modified to support using as the retriever.
5
63
317
This now means you can load:. - PDFs.- PowerPoints.- HTML.- Word Docs.- Images (.jpg, .png).- Email documents. Into a format @LangChainAI can work with. all with a single command.
Before you can use LangChain with your data, you first need to clean it up. That's where an integration with @UnstructuredIO comes in. Blog Post: We'll use Unstructured to power a lot of our document loaders:
8
30
316
ChatGPT is not amazing at following instructions on how to output messages in a specific format. This is leading to a lot of `Could not parse LLM output` errors when trying to use @LangChainAI agents. We recently added an agent with more strict output formatting to fix this. 👇.
12
33
317
It's out! LangChain v0.1.0 comes out with an improved package architecture for stability and production readiness, as well a focus on:. 👀Observability.↔️Integrations.🔗Composability.🎏Streaming.🧱Output Parsing.🔎Retrieval.🤖Agents.
🦜🚀LangChain v0.1.0. After a year of development, we've released LangChain v0.1.0. Read the full blog here: After talking with the users and developers, we released this stable version and focused the library on a few core areas:. 👀Observability:
28
49
308
🧠I wrote some thoughts on memory for agents!. We released a bunch of new functionality for memory in LangGraph, and in doing so we thought hard about what memory actually means, and was is useful today. Some highlights 👇. 🛃Memory is application specific. The best memory today
13
51
316
⭐️Multi-Action Agent w/ @OpenAI Functions⭐️. OpenAI functions allows the LLM to select a single tool to use. but often times it can be more efficient to have the LLM select MULTIPLE tools to use (if tasks can be done in parallel). Enter. the `openai-multi-functions` agent👇
14
57
313
Lots of great small features added over the past week. My favorites: improvements to OpenAI tracking to allow for cost tracking (@tim_asp) + number of successful requests (Jon Page)
In the midst of many big releases last week, there was also a lot of smaller improvements/bug fixes that got in. Although not as flashy as big features, these are integral in making sure LangChain is as usable as possible. Here's a list:. 👇.
9
32
312
"12 Factor Agents - Principles for building reliable LLM applications". Great piece from @dexhorthy. A lot of the principles really resonate with LangGraph ethos. Would people be interested in a blog on how LangGraph enables these?. #2: Own your prompts (no hidden prompts in
14
46
316
A year ago, there wasn't enough "tinkering" happening (h/t @natfriedman). 🦜🔗That's why we started LangChain. Now, there aren't enough applications making their way from prototype to production. 🦜🛠️That's why we're launching LangSmith.
🦜🛠️ Introducing LangSmith 🦜🔗. A unified platform to help developers debug, test, evaluate, and monitor their LLM applications. Integrates seamlessly with LangChain, but doesn't require it.
15
35
308