Founder and CTO - Daylight Security. Cybersecurity, AI and everything in between

Israel
Joined July 2010
Pragmatic AI Hacks #1: Use Claude + Google Calendar. Instantly get personal assistant-level scheduling with multi-timezone support, semantic prioritization rules (e.g. 'Always prioritize customer meetings over weekly meetings if no other valid time slot found'), and more.
3
251
TDD in 2025 == Tab Driven Development
1
8
468
I built this E2E SaaS in 1 day of real work with ZERO front-end experience ๐Ÿคฏ Using @cursor_ai, @v0 and pure English. Its called dandi.cloud powered by @langchain JS ๐Ÿฆœ๐Ÿ”— and it outputs curated LLM digested information and insights about open source repositories I don't know JS, not HTML, not CSS, just natural language and debugging. Features: โ—† RESTful API endpoints โ—† Authentication โ—† API Key Management โ—† Rate limitng โ—† AI Integration with @langchain Tech Stack (TBH Cursor did all the work) @vercel @nextauthjs @Shadcnkit @tailwindcss @v0 @langchain Code is open-sourced - you can clone it and MICRO SAAS Season is ON!!!! github.com/emarco177/dandi #opensource #buildinpublic
9
12
2
69
16,244
Amazing!
What can an 8-year-old build in 45 minutes with the assistance of AI? My daughter has been learning to code with @cursor_ai and it's mind-blowing๐Ÿคฏ Here are highlights from her second coding session. In 45 minutes she built a chatbot powered by @CloudflareDev Workers AI ๐Ÿ‘€
3
252
Coding with @cursor_ai for the past few days, It's amazing and definitely change the way I code
9
220
The Unspoken Truth About Autonomous Agents: Why LangGraph ๐Ÿฆœ๐Ÿ•ธ๏ธ by @langchain Will Dominate in 2024. Fully autonomous agents๐Ÿค– arenโ€™t really working because we give too much freedom to the LLM and rely on them to handle it. They can't. This leads to unrealistic expectations and disappointing results. ๐Ÿ‘‡ Prediction for 2024: Most production served agentic application will be implemented with LangGraph by @langchain ๐Ÿฆœ๏ธ๐Ÿ”— . LangGraph offers a solution by scoping the freedom of LLMs, allowing developers to create controlled, deterministic flows (with cycles!). These flows can be navigated by the LLM, leveraging its reasoning capabilities while maintaining developer control๐Ÿ‘ฉ๐Ÿปโ€๐Ÿ’ป. This balance is key to building robust, maintainable WORKING agentic architectures. My favorite LangGraph examples: GPT Researcher by @assaf_elovic : A great example illustrating how LangGraph can implement sophisticated multi-agent workflows, showcasing LangGraph's practical, production-ready potential yielding quality results. github.com/assafelovic/gpt-rโ€ฆ Advanced RAG Architectures from @MistralAI & @langchain : by @sophiamyang and @RLanceMartin contributions in the LangChain Mistral cookbook. They demonstrate cool RAG techniques like corrective RAG, adaptive RAG, and self-reflective RAG, written in an elegant LangGraph flow. github.com/mistralai/cookbooโ€ฆ BTW both examples use @tavilyai which is gaining momentum as well๐Ÿš€ and connects our LLMs to real-world data. The great thing here is that the results are digestible by our genai app, and the purpose of this service is to generate results that are downstreamable into our genai app. tavily.com/ In the video below, I elaborte on autonoums agents today and explain how LangGraph provides developers with control while still leveraging the powerful reasoning capabilities of LLMs, making it the most elegant solution for flow engineering today. youtube.com/watch?v=VXHNke_uโ€ฆ My prediction? LangGraph๐Ÿฆœ๐Ÿ•ธ๏ธ by @langchain will THE major agent player in 2024, enabling developers to build advanced, production-oriented agents with ease.
5
31
1
153
21,087
I just built an Insanely Complex RAG Flow with @langchain's LangGraph โ€“ You Won't Believe How Easy It Is I've been working on an open source git repo for advanced RAG flows with @langchain 's LangGraph๐Ÿฆœ๐Ÿ•ธ๏ธ, heavily inspired by the LangChain Cookbook by @RLanceMartin and @sophiamyang ! This repo not only implements Corrective RAG, Adaptive RAG, and Self-RAG with LangGraph but also focuses on structuring the code for maintainability, testing & clean code. ๐ŸŒŸ We leverage LangGraph to build an advanced RAG flow using ideas from 3 papers: Corrective-RAG (CRAG): Self-grading on retrieved documents and web-search fallback. Self-RAG: Self-grading on generations for hallucinations. Adaptive RAG: Routes queries based on complexity. Explore the repo here: github.com/emarco177/langgapโ€ฆ Youtube walkthrough here: youtube.com/watch?v=NZbgduKlโ€ฆ Let me know what you think! ๐Ÿ’ฌ
8
41
1
252
30,135
Had an amazing time on the main stage advocating for GenAI Agents at the @googlecloud Summit! 2024 is truly the year of the agents๐Ÿค– I loved championing open source on Google Cloud and showcasing @langchain ๐Ÿฆœ๏ธ๐Ÿ”— , specifically LangGraph, to implement complex agentic flows. LangGraph๐Ÿฆœ๐Ÿ•ธ๏ธ is a cutting-edge flow engineering implementation that simplifies the creation and management of complex agentic flows. Deploying GenAI agents on Google Cloud has never been more exciting, offering robust infrastructure, scalability, and seamless integration with innovative AI tools. It was a pleasure hosting the one and only Oded Shvartz from @torq_io , who shared their groundbreaking HyperSOCโ„ขย  work on their state-of-the-art autonomous agent, Socrates, which is revolutionizing SOC teams all over the world.
1
9
1
30
16,908
ืขื•ื“ ืจื’ืข Adagio for Strings ื—ื•ื’ื’ 20 ืฉื ื” (!), ืžื“ื”ื™ื ืœืจืื•ืช ืื™ืš ื”ื•ื ืขื“ื™ื™ืŸ ืจืœื•ื•ื ื˜ื™, ืžืฆืžืจืจ ื•ืžืจื™ื ืืช ื”ืงื”ืœ ืขื“ ื”ื™ื•ื youtu.be/KO7cYLLFbLw?si=n84Fโ€ฆ
1
1
205
๐ŸŽ—๏ธ๐Ÿ’”
878
1,737
140
10,003
277,366
Eldad Rudich retweeted
ื—ื–ืจืชื™ ืฉืœืฉื•ื ืœื˜ื™ืคื•ืœ. ื˜ื™ืคื•ืœ ืคืกื™ื›ื•ืœื•ื’ื™. ื—ืฉื‘ืชื™, ืฉื›ืœ ืขื•ื“ ื”ืžืœื—ืžื” ืžืชื ื”ืœืช ื•ืื ื™ ืขื•ื‘ื“, ืื– ืื™ืŸ ื‘ืืžืช ืฆื•ืจืš. ื”ื›ืœ ื‘ืกื“ืจ. ืื‘ืœ ื‘ื›ืœ ืœื™ืœื”, ื›ืฉื”ื—ื•ืฉืš ื™ื•ืจื“ - ืื ื™ ืžืชืคืจืง. ื•ื‘ื›ืœ ืœื™ืœื” ืฉืขื•ื‘ืจ ื–ื” ืžื—ืžื™ืจ. ื‘ื™ื•ืžื™ื™ื ื”ืื—ืจื•ื ื™ื ืกื™ื•ื˜. ื ืจื“ื ืœืฉืขื”, ื•ืื– ืžืชืขื•ืจืจ ืžื—ืœื•ื ืจืข ืขื ื”ืชืงืฃ ื—ืจื“ื” ื ื•ืจืื™. ืกื™ื•ื˜ ืฉื”ื™ื” ื”ืžืฆื™ืื•ืช ืฉืœ ืžืื•ืช ืžื”ื—ื‘ืจื•ืช ื•ื”ื—ื‘ืจื™ื ืฉืœื™ ื‘ืฉื‘ืช ืฉืขื‘ืจื”. ืกื™ืคื•ืจื™ื ืฉืฉืžืขืชื™ ื•ืžืฉืชืœื˜ื™ื ืขืœื™. ืงื ืžื–ื™ืข, ืžื‘ื•ืขืช. ื•ืื– ืฉืขื” ืืจื•ื›ื” ืฉื”ื›ืœ ืฉื—ื•ืจ ื•ืืชื” ืžืจื’ื™ืฉ ืกื’ื•ืจ ื‘ืชื•ืš ืืจื‘ืข ืงื™ืจื•ืช ื•ืœื ืžืฆืœื™ื— ืœื‘ืจื•ื— ื”ื—ื•ืฆื”. ืชื—ื•ืฉื” ืฉื–ื” ื”ืกื•ืฃ. ืกื•ืฃ ื”ืขื•ืœื. ื•ื–ื” ืœื ืžืฉื”ื• ื—ื“ืฉ ืขื‘ื•ืจื™. ื”ืชืงืคื™ ื—ืจื“ื” - ื–ื” ืžืฉื”ื• ืฉืื ื™ ืžืชืžื•ื“ื“ ืืชื ื›ื‘ืจ 5 ืฉื ื™ื. ื‘ื“ืจืš ื›ืœืœ ืื ื™ ืžื ืฆื— ืื•ืชื. ืื‘ืœ ื”ืคืขื ืœื ืžืฆืœื™ื— ืœื”ืชืžื•ื“ื“. ืืคื™ืœื• ืคืจืง ืฉืœ ื—ื‘ืจื™ื ( ื”ืกื“ืจื” ืฉืื ื™ ื”ื›ื™ ืื•ื”ื‘, ื”ื‘ืจื™ื—ื” ืฉืœื™ ืžื”ื”ืชืงืฃ) ืœื ืขื•ื–ืจ. ื”ื•ื ืžื ืฆื— ืื•ืชื™. ื”ื”ืชืงืฃ. ื•ืฉืœืฉื•ื, ื”ื—ืœื˜ืชื™ ืฉื–ื”ื•- ื—ื–ืจืชื™ ืœื“ื‘ืจ ืขื ืื™ืฉ ืžืงืฆื•ืข ื•ื–ื” ืขื•ืฉื” ืคืœืื™ื. ืคื™ืจืงื ื• ื‘ื™ื—ื“ ืืช ื™ื•ื ืฉื‘ืช ื”ืฉื—ื•ืจ. ืืช ื”ืคื—ื“ ืฉืœื ื•, ืฉืœื™, ืฉืœ ื”ื™ืœื“ื™ื ืฉืœื™ ื•ืฉืœ ืืฉืชื™ ืฉื™ื•ืฉื‘ื™ื ื‘ืžืžื“. ืืช ื”ื”ื•ื“ืขื•ืช ืฉืงื™ื‘ืœืชื™ ืžื”ื—ื‘ืจื™ื ืฉืฉื•ืžืขื™ื ืืช ื”ืžื—ื‘ืœื™ื ื‘ืชื•ืš ื”ื‘ื™ืช ื•ื–ื•ืขืงื™ื ืœืขื–ืจื”, ืืช ืืœื” ืฉื”ืฆืœื—ืชื™ ืœื›ื•ื•ืŸ ืืœื™ื”ื ื›ื•ื—ื•ืช ื•ืฉืจื“ื•, ื•ืืช ืืœื” ืฉืœื - ื•ื”ื ื›ื‘ืจ ืœื ืืชื ื•. ื—ื‘ืจื•ืช ื•ื—ื‘ืจื™ื ืฉืื ื™ ืžืจื’ื™ืฉ ื›ื‘ืจ 12 ื™ื•ื, ืฉื™ืฉ ืœื™ ื—ืœืง ื‘ื–ื” ืฉื”ื ืœื ืฉืจื“ื•. ืคื™ืจืงื ื• ื‘ื™ื—ื“ ืืช ื”ืฉืขื•ืช ื”ืจืืฉื•ื ื•ืช, ืžืžื˜ื—ื™ ื”ืจืงื˜ื•ืช ืฉื”ืงืคื™ืฆื• ืื•ืชื™ ืžื”ืžื™ื˜ื” ื•ืขื“ ืฉื™ื—ืช ื”ื˜ืœืคื•ืŸ ืฉื‘ืžื”ืœื›ื” ืืชื” ืžื‘ื™ืŸ ื‘ืคืขื ื”ืจืืฉื•ื ื” ืฉื”ืžื—ื‘ืœื™ื ื›ื•ื‘ืฉื™ื ืœืš ืืช ื”ืขื•ื˜ืฃ ืฉืœืš. ื•ื’ื ืืช ื”ืจื’ืขื™ื ืฉืื ื™ ืขื•ืฆืจ ืœืจื’ืข ืืช ื”ืฉื™ื“ื•ืจ ื›ื“ื™ ืœืืกื•ืฃ ืžื™ื“ืข ื•ืื– ืื™ืœืช ืืฉืชื™ ื‘ืื” ืœื“ืจื•ืฉ ื‘ืฉืœื•ืžื™ ื•ืื ื™ ืžืชืคืจืง. ื‘ื•ื›ื” ื›ืžื• ื™ืœื“. ื—ืกืจ ืื•ื ื™ื. ื™ื•ื“ืข ืฉื”ืื ืฉื™ื ืฉื“ื™ื‘ืจืชื™ ืืชื ืœืคื ื™ ืจื’ืข, ืœื ื™ืฉืจื“ื• ื•ืื™ืŸ ืœืš ืžื” ืœืขืฉื•ืช. ืžืชืงืฉืจ ืœืื ืฉื™ ื‘ื™ื˜ื—ื•ืŸ ืฉืื ื™ ืžื›ื™ืจ, ืœืคืจืืžื“ื™ืงื™ื ืฉืœ ืžื“ื, ืœื’ื•ืจืžื™ื ื”ื›ื™ ื‘ื›ื™ืจื™ื ื‘ืžืขืจื›ืช ื•ื›ื•ืœื, ื›ื•ืœื ื• ื—ืกืจื™ ืื•ื ื™ื. ืคื™ืจืงื ื• ืืช ื–ื” ื™ื—ื“. ื•ืื—ืจื™ ื–ื” ืื ื™ ืฉื•ื‘ ืขื•ืžื“ ืขืœ ื”ืจื’ืœื™ื™ื. ื—ื–ืง ื™ื•ืชืจ (ืื‘ืœ ืžืช ืžืคื—ื“ ืžื”ืœื™ืœื”). ื•ืื ื™ ืžืฉืชืฃ ืขื›ืฉื™ื• ื›ื™ ืื ื™ ืฉื•ืžืข ืฉืชื•ืฉื‘ื™ื ื‘ืขื•ื˜ืฃ, ื’ื‘ืจื™ื, ื ืฉื™ื ื•ื™ืœื“ื™ื, ืžืชื‘ื™ื™ืฉื™ื ืœืœื›ืช ืœื˜ื™ืคื•ืœ. ืืœ ืชืชื‘ื™ื™ืฉื•. ื–ื” ืžืฆื™ืœ ื—ื™ื™ื. ื–ื” ื™ืฆื™ืœ ืืชื›ื ื•ื™ืฆื™ืœ ืืช ื”ื™ืœื“ื™ื ืฉืœื›ื. ื›ื›ื” ืืชื ืชื“ืขื• ืœืชื•ื•ืš ืœื”ื ืืช ื”ื–ื•ื•ืขื•ืช ืฉื—ื•ื•ื™ื ื• ื›ืืŸ. ื•ืชืขื–ืจื• ืœื”ื ืœืขืžื•ื“ ืขืœ ื”ืจื’ืœื™ื™ื. ืชืคืจืงื• ืืช ืžื” ืฉื™ื•ืฉื‘ ืœื›ื ืขืœ ื”ืœื‘. ืชืคืจืงื• ืืช ืžื” ืฉื—ื•ื•ื™ืชื ื‘ืื•ืชื• ื™ื•ื ื ื•ืจื. ืชืชืคืจืงื•. ื”ืžื˜ืคืœื™ื ื•ื”ืžื˜ืคืœื•ืช ื”ื ื”ืžืœืื›ื™ื ื•ื”ืžืœืื›ื™ื•ืช ืฉื”ืฆื™ืœื• ืื•ืชื™ ื‘ืขื‘ืจ ื•ื”ื ื’ื ื™ืฆื™ืœื• ืืชื›ื๐Ÿ’” ืœื›ื• ืœื˜ื™ืคื•ืœ.
1,357
1,350
221
15,438
599,646
43
202
22
1,564
134,862
RT @Israel: Breaking: IDF Spokesperson From the analysis of the operational systems of the IDF, an enemy rocket barrage was carried outโ€ฆ
Eldad Rudich retweeted
Kidnapping children is despicable and wholly unjustifiable. For obvious reasons, this picture has hit home with me. May Noya and all hostages taken by Hamas be returned soon, safely, to their families.
This beautiful 12 year old girl with autism was kidnapped from her home by Hamas terrorists and was taken to Gaza. Noya, is sensitive, kind, funny and a massive Harry Potter fan. @jk_rowling can you help us get her story out? Share this and help us bring Noya home โ™ฅ๏ธ #HamasisISIS
13,389
18,931
3,258
131,011
27,549,892
19 ื‘ืื•ืงื˜ื•ื‘ืจ 1987. ื™ื•ื ืฉื ื™ ื”ืฉื—ื•ืจ. ืฉื•ืง ื”ืžื ื™ื•ืช ืงืจืก ื‘ื™ื•ืชืจ ืž-20% ื‘ื™ื•ื ืื—ื“. ื ืืกื™ื ื˜ืืœื‘ ื™ืฉื‘ ื“ืจื•ืš ื‘ืงืฆื” ื”ื›ืกื ืฉืœื• ื‘ืจืฆืคืช ื—ื“ืจ ื”ืžืกื—ืจ ืฉืœ ื‘ื ืง ื”ื”ืฉืงืขื•ืช ืคื™ืจืกื˜ ื‘ื•ืกื˜ื•ืŸ. ืกื‘ื™ื‘ื• ื”ื™ื™ืชื” ืžื”ื•ืžื”. ืื‘ืœ ืœื˜ืืœื‘ ื“ื•ื•ืงื ื”ืœืš ื‘ืกื“ืจ ื‘ืื•ืชื• ื™ื•ื. ื•ื”ื•ื ืœื ื™ื“ืข ืฉื–ื” ื™ื”ื™ื” ื”ื™ื•ื ืฉื™ื‘ื ื” ืืช ื”ืงืจื™ื™ืจื” ืฉืœื•. 1/
32
20
11
662
139,976
1/17๐ŸงตDemystifying LLM memory๐Ÿง  mega thread featuring @langchain ๐Ÿฆœ๐Ÿ”— In this thread I will cover the most popular real-world approaches for integrating memory to our GenAI applications ๐Ÿค–
11
72
4
394
94,904
Eldad Rudich retweeted
Check out my new blog post about how we shrunk the storage size of our CloudSQL Postgres database by migrating the data with no downtime using pg_logical extension. @googlecloud @PostgreSQL kostyay.medium.com/zero-downโ€ฆ
1
7
218
Eldad Rudich retweeted
Torq recommends CircleCI users rotate their secrets immediately to mitigate the effects of yesterdayโ€™s large-scale breach. Torq is offering a free account and assistance to any organization looking to automate and orchestrate the process. Learn more: torq.io/blog/circleci-breachโ€ฆ
2
5
656