DAGWorks Inc.

Really excited to talk about building out #AIAgents reliably with #Burr at PyData Global next week! https://lnkd.in/gafKvBM9 If you haven't already, make sure to register -- great roster of speakers!

A #Burr user shared this video https://lnkd.in/gie9N6TG with me on the "problem with frameworks". Worth a watch. I ??% agree with the take home. The "coupling problem" is the reason why people graduate out of frameworks like LangChain & LlamaIndex. You become overly coupled and then find yourself overtime asking why am I using this? It's not that the framework authors did this maliciously, it's just that they optimized for their needs, in this case for POCs, which is not the same as long lived maintainable code. That's why we see people either build their own frameworks, or they hopefully discover #hamilton & #burr and realize that there are frameworks that allow "loose coupling" that don't get in your way (as a platform person and framework author we optimize for maintainable code - and the user who shared the video agrees and is why they use Burr). A good example of this is FastAPI - very easy to not couple your business logic and iterate quickly no matter the age of the code. So how can you tell where a framework lies on the spectrum of tight coupling versus not? Some tests: (1) How many "objects" do I need to use from the framework to couple the logic I want to happen? The fewer the better. E.g. with #hamilton by default it's 0. With #burr it's just 1 and that's due to state for edge transitions, but that's easy to decouple from. (2) How many framework imports do you need to get something to run? The less the better - consider #burr vs LangChain using #langgraph (see image). If you're feeling framework pains, or are curious on how it can be done right, checkout the links in the comments, or comment yourself/tag someone who you think would like this post. #python #opensource #genai #llmops

Happy Friday! Here's what happened: > #Hamilton release highlights: User contributed async support for @pipe (thanks Jernej Frank!) + various small fixes > #Burr release highlights: ?? || . Parallelism V1 is out! This is ??. It's a highly asked for construct. You could do it before, but now it's simpler to do! Excited for people to build more parallel actions / sub-agents! More in the newsletter. > Office Hours & Meet ups for Hamilton & Burr. Sign up, come and ask questions. We'd love to see you there. Details in the newsletter. > Free Lightning lesson on #GenAI from first principles for SWEs, DS, and MLEs with Hugo Bowne-Anderson on Maven on Tuesday - sign up! Link in the newsletter. > Blog post: Flashcard generator with #Instructor + #Burr. This is a fun post showing what you can do with GenAI and easily and quickly with Instructor & Burr! > In the wild: a few mentions on LinkedIn for both Hamilton & Burr Thanks to Eric Brichetto, Dr. Milan Jelisav?i?, & Unwind AI for the shoutouts!

Are you struggling to figure out how to build out a process to deliver on #GenAI? Sign up for Stefan Krawczyk and Hugo Bowne-Anderson's free session that will help you understand the #SDLC you'll need to build out to ship production grade experiences.

Building AI agents is hard But tracking how they think shouldn't be. Meet Burr - an opensource Python library that lets you see inside your AI agent's mind. Here's what makes it special: → Visual debugging in real-time → Built for production from day one → Works with any Python code, not just LLMs → Simple building blocks, powerful results → Complete toolkit for deployment The magic? State machines. Instead of adding endless print statements or digging through logs, Burr shows you exactly how your agent thinks through an interactive graph. Want to know why your chatbot gave that response? Just click on any node to see the exact state and data at that moment. Need to add safety checks? Draw the logic flow and Burr handles the rest. Building features like: ? Conversation history ? Complex decision trees ? Streaming responses ? State persistence ? Safety guardrails All become visual building blocks you can see and debug. The best part? If you can write a Python function, you can build an agent. No complex abstractions. No vendor lock-in. No black boxes. Just clear, debuggable, production-ready AI agents. If you find this useful, Like ?? and share ?? this post with your network Don't forget to follow Unwind AI for more such AI tips and tutorials.

Excited to share one of my latest projects: TSGPT (Trust & Safety GPT), a multilingual content moderation system inspired by industry-leading tools used at platforms like Meta! Over the summer, I dove deep into building a content moderation and detection system from scratch, using a combination of large language models (LLMs), semantic similarity, graph databases, Diffbot enrichment APIs, and policy-driven enforcement. The goal was to create a tool capable of identifying and responding to extremist content across multiple languages and scripts, while balancing platform safety and user expression. Key Features of TSGPT: ? Multilingual Extremism Classification: Detects and classifies extremist content across 8+ languages ? Dangerous Entity Glorification Detection: Uses a Neo4j knowledge graph of dangerous organizations with aspect-based sentiment analysis to detect glorification or material support of hate groups & terror entities, while protecting neutral or condemnatory speech acts about those entities ? Nuanced Policy Enforcement: Implements context-based actions that can range from content warnings to visibility restrictions. ? Scalable, Modular Architecture: Built using Hamilton DAG Framework from Stefan Krawczyk and DAGWorks Inc., self-hosted LLMs, and Neo4j, ensuring adaptability for future enhancements. This project was the capstone of my recent professional development courses, Building LLM Applications From Scratch and Advanced LLM Application Building, both taught by Hamza Farooq through Maven. Through these courses, I learned invaluable skills in building scalable, production-ready LLM applications—skills that have accelerated my career trajectory, attracted new professional opportunities, and equipped me with the technical expertise to tackle real-world Trust & Safety challenges head-on. ?? Check out the full video presentation of TSGPT here: https://lnkd.in/gXGhrESB Thank you to Stefan Krawczyk for generously motivating me to push this to fruition and thank you to Maven for providing an exceptional learning experience. If you're looking to level up your expertise in LLM application engineering, I highly recommend these courses.

I've been working on integrating #burr into Wingman SQL Assistant, a tool designed to transform natural language into SQL for simplified data access. This project is a step toward incorporating these capabilities into the main Salesteq engine, making data interactions more intuitive. Explore it here and see how AI can change the way you work with data: app.salesteq.ai/chat We're steadily advancing in the journey of generative AI for enterprise tech.

System Prompt Improvement Using Dialog Engineering And Utilizing The Burr Framework By DAGWorks Inc. https://lnkd.in/eJSX8-7X

Happy Thursday! Here's the news: > #Hamilton release highlights: - in-memory cache store. #1 feature request since releasing caching. You can now use this feature in particular for notebook iteration with Hamilton. So keep everything in memory == fast iteration, and then when you're done, checkpoint it to disk. > #Burr release highlights: - want to have parallel agents? See our release candidate. Details in the post. > #GenAI Education - Our workshop at MLOps World & Generative AI World Summit 2024 is next week. We have free passes to give away to our workshop for SWEs - details in the newsletter. - I'm running a Maven course with Hugo Bowne-Anderson & we need input! So if you're interested how to better build and ship GenAI products, we'd love to understand what would be interesting to you. Link in the newsletter to the survey. > Blog post: - Thanks to Thierry Jean for a post on using #haystack (deepset) with #burr. Post is titled: Build LLM agents faster with Haystack + Burr! Link in the newsletter. > In the wild: - Both Hamilton & Burr were at the #dataforAI event! > From our friends: - dltHub event with Foundation Capital. > Office Hours & Meet ups for Hamilton & Burr. - link in the newsletter.

Hey Friends, great opportunity here ?? ??? Know a Software Engineer interested in breaking into AI (building GenAI powered apps for scale)? We’ve got 3 FREE tickets up for grabs for an exclusive two-day, hands-on workshop at the MLOps World / Generative AI World Summit in Austin ?? Join "Building GenAI-Powered Apps: A Workshop for Software Engineers" led by industry experts Hugo Bowne-Anderson and Stefan Krawczyk. It’s packed with real-world insights and hands-on guidance to help you dive into generative AI development. ?? Tag a SWE friend or comment if you’re interested—we’ll randomly select 3 winners! Best of luck! Let’s get more SWEs building GenAI! ???