Empowering the InfoSec Community through Open Source projects and collaboration! https://teapawt.net/t.co/T9YKVakZ9o
Joined January 2020
- Tweets150
- Following5
- Followers4.6K
- Likes212
Open Threat Research retweeted
β‘οΈ In an effort to augment existing @OTR_Community projects with AI, @Cyb3rPandaH and I are extending the @HunterPlaybook using Agent Skills π€ to describe known workflows and leverage progressive disclosure, so agents access the right knowledge at the right time.
πΉ The hunt lifecycle (plan β‘οΈ execute β‘οΈ report) needs structure, and we recently took some time to show how the planning stage can be expressed as Agent Skills using @code and GitHub Copilot, together with @tavilyai and @Microsoft Sentinel Data Lake MCP Servers π₯.
Blog post link in the comments.
Open Threat Research retweeted
Awesome talk from @Cyb3rWard0g on AI Agents... this for me is the immediate future of LLM's and it's so exciting! youtube.com/watch?v=zoAPS1gsβ¦
Open Threat Research retweeted
Thank you for the opportunity @x33fcon ! One of the best conferences of the year! I appreciate all the support πβ€οΈ
Open Threat Research retweeted
β¨ What an amazing week teaching at @x33fcon with @Cyb3rPandaH ! π We dove into AI Agents for Security Operations from the ground up and enjoyed all the great questions from such curious minds!
π Thanks for making it an unforgettable experience. Canβt wait to see what you build next! #poland #gdynia #x33fcon @OTR_Community
Open Threat Research retweeted
Itβs amazing to see @daprdev agents being featured in a hackathon at MS Build!!! β€οΈππ₯ LG!!! Canβt wait to see what people build with it and the feature requests π @diagridio
Join the Dapr team for a hands-on hack during MS Build in Seattle!
Dapr co-creators Mark Fussell & Yaron Schneider will lead a casual, hands-on hackathon featuring Dapr Workflows & Dapr Agents.
We will demo these new functionalities, then be available while you snack, drink, and code.
π Place:Β Kells Irish Restaurant & Pub
ποΈΒ May 20
π§Β Kickoff ~6 PM - Dapr Agents & Workflows demo.
Hang out with the Diagrid team and other MS Build attendees
buff.ly/nZsOiI4
Open Threat Research retweeted
@daprdev agents demos ππ₯β€οΈ happening now!
Open Threat Research retweeted
π That moment when you realize that what started as a weekend curiosity to learn more about agentic systems π€ has become a project others find useful and in innovative β€οΈ Thank you @diagridio and @bibryam for sharing! @daprdev Agents!
π New on the Diagrid Blog: Building Effective Dapr Agents
@bibryam unpacks the spectrum of agentic behavior and how to design intelligent, production-grade agents using Dapr Agents.
He shares 8 practical patternsβfrom simple prompt chains to fully autonomous agentsβusing real-world examples and production-ready code.
Learn why predictable workflows often outperform fully autonomous agents in enterprise settingsβand how Dapr helps you strike the perfect balance.
buff.ly/WUbAI18
#Dapr #AgenticAI #PlatformEngineering #LLM #Microservices #Diagrid
Open Threat Research retweeted
β€οΈ Happy to share my journey in creating an AI Agentic Workflow Engine with @daprdev for developer-friendly, code-first workflows, agent orchestration, state management, and communication.
blog.openthreatresearch.com/β¦
β‘οΈ Building my own AI Agentic framework was an incredibly rewarding experience, leading to the creation of project Floki πβπ¨!
π Iβm also very excited to announce that Iβve decided to donate the Floki codebase to the @daprdev project and am partnering with the Dapr community to evolve it further. Today, Dapr also launched "Dapr Agents" π€ the evolution of Floki, and Iβm very proud to see it join the Dapr community β€οΈ! #opensource #AI #Dapr @OTR_Community
Open Threat Research retweeted
π Finishing the long weekend by completing Unit 1: AI Agents Fundamentals from @huggingface π. I learn best by coding, so Iβm sharing my notebooks that help me cover some of the fundamentals in each unit! π€
π Repo: github.com/Cyb3rWard0g/HF-Agβ¦
π My Unit 1 notes cover:
β
Downloading & setting up SmolLM2 locally
β
Performing next token prediction
β
Defining tools for agents
β
Writing ToolCalling & ReAct agents from scratch with transformers π & SmolLM2 π€
π Excited to dive into the next units and share more! Follow along for more! β¨ @OTR_Community
Open Threat Research retweeted
β¨ Early registration for #BHUSA 2025 trainings is now open π Whether you're defining POCs, MVPs, or requirements for agentic systems, my class will give you hands-on experience and dive into core components of agentic workflows to build adaptive, autonomous systems for any security challenge π
π Register here: blackhat.com/us-25/training/β¦
π£ANNOUNCEMENT: Registration for #BHUSA 2025 is now live! Returning to the Mandalay Bay Convention Center in Las Vegas, August 2-7. Join us for a six day program including #cybersecurity Trainings, Briefings, Summits, Arsenal, a robust business hall, networking & more! View pass options: bit.ly/4hhmGqS
Open Threat Research retweeted
π’ Just a reminder β‘οΈ There is still time to apply for these exciting roles on my team! Applications are accepted until January 12, 2025 ποΈ!
β¨Weβre looking to proactively explore AI in offensive security operations, improve our simulation environments (agentic playgrounds), and continue researching and enabling autonomy to defend Microsoft! π
β¨Before the year ends, Iβm super excited to share that my team continues to grow! π€
As a firm believer that offense informs defense and vice versa π, our Autonomous Defense and Protection Team (ADAPT) is proactively exploring how AI can enrich offensive security operations β‘οΈ.
Weβre also continuing to invest in elevating our simulation environment infrastructure to support agentic playgrounds and simulate attack scenarios for both training and evaluation of AI systems βοΈ. Last but not least, we continue to invest in enabling autonomy to defend our organization π¦Ύπ‘οΈ.
π’Iβm hiring for 4β£ exciting roles:
π₯ Principal Security Researcher: An experienced Red Team Operator with engineering expertise who will collaborate with applied scientists to leverage LLMs and agentic systems to identify and operationalize adversary techniques. π jobs.careers.microsoft.com/gβ¦
π€ Principal Offensive AI Scientist: Drive innovation using AI to enhance offensive security capabilities while partnering with experienced offensive security researchers. π jobs.careers.microsoft.com/gβ¦
βοΈ Principal Security Engineer: Build secure, scalable environments for attack simulations, supporting agentic playgrounds and realistic adversary scenarios. π jobs.careers.microsoft.com/gβ¦
π€ Principal Applied Scientist: Design AI-driven workflows to enable autonomy and tackle complex security challenges. π jobs.careers.microsoft.com/gβ¦
Positions are 100% remote and US-based (potential expansion for the right candidate). Microsoft will accept applications for the role until January 12, 2025, ποΈ. #Cybersecurity #LLM #AIAgents #GenAI #GenerativeAI #AI #hiring #redteam #blueteam
Open Threat Research retweeted
β¨Before the year ends, Iβm super excited to share that my team continues to grow! π€
As a firm believer that offense informs defense and vice versa π, our Autonomous Defense and Protection Team (ADAPT) is proactively exploring how AI can enrich offensive security operations β‘οΈ.
Weβre also continuing to invest in elevating our simulation environment infrastructure to support agentic playgrounds and simulate attack scenarios for both training and evaluation of AI systems βοΈ. Last but not least, we continue to invest in enabling autonomy to defend our organization π¦Ύπ‘οΈ.
π’Iβm hiring for 4β£ exciting roles:
π₯ Principal Security Researcher: An experienced Red Team Operator with engineering expertise who will collaborate with applied scientists to leverage LLMs and agentic systems to identify and operationalize adversary techniques. π jobs.careers.microsoft.com/gβ¦
π€ Principal Offensive AI Scientist: Drive innovation using AI to enhance offensive security capabilities while partnering with experienced offensive security researchers. π jobs.careers.microsoft.com/gβ¦
βοΈ Principal Security Engineer: Build secure, scalable environments for attack simulations, supporting agentic playgrounds and realistic adversary scenarios. π jobs.careers.microsoft.com/gβ¦
π€ Principal Applied Scientist: Design AI-driven workflows to enable autonomy and tackle complex security challenges. π jobs.careers.microsoft.com/gβ¦
Positions are 100% remote and US-based (potential expansion for the right candidate). Microsoft will accept applications for the role until January 12, 2025, ποΈ. #Cybersecurity #LLM #AIAgents #GenAI #GenerativeAI #AI #hiring #redteam #blueteam
Open Threat Research retweeted
πβ‘οΈHappy to share the first release of Floki! π€
π₯ Repo: github.com/Cyb3rWard0g/flokiβ¦
A framework Iβve been working on to dive deeper into multi-step agentic workflows, event-driven orchestration of multi-agent systems, and shared communication, all powered by @daprdev β€οΈ
π― Service Invocation: Built-in service discovery and error handling for agent communication.
β‘οΈ Pub/Sub: Real-time, event-driven collaboration via a shared message bus.
π Workflow API: Autonomy and deterministic flows with LLM-based decision-making.
π§ State Management: Agents retain context across interactions.
π€ Actors: Agents as virtual actors that handle messages sequentially for scalability and simplicity.
Open Threat Research retweeted
Since starting my Cybersecurity career with a background in AI, Iβve dreamed of collaborating with a team passionate about both fields π
Today my dream becomes a reality ππ½ I am excited to join @BlueVoyant as an AI Researcher π€© and I canβt wait to take on new challenges π€π
Open Threat Research retweeted
π£ As part of the Autonomous Defense & Protection Team (ADAPT), my team is growing and taking shape with recent hires, and Iβm excited to announce a new opportunity to join us!
π₯Apply here β‘οΈ jobs.careers.microsoft.com/gβ¦
π Iβm hiring a Principal Applied Scientist to support our current research efforts in enabling autonomy via LLM-based AI agents during investigations and hunting operations, with the main goal of defending and protecting Microsoft.
β
Dive deep into knowledge graph research, integrating these concepts within LLM and multi-agent workflows to equip AI agents with the right information and context for critical decisions in security operations.
β
Generate synthetic data and environments, creating realistic attack scenarios to serve as the foundation for training autonomous agents.
β
Help build frameworks that support the development of an agentic playground for red vs. blue operations, enhancing the evaluation of agent capabilities across security scenarios.
Positions are 100% remote and US-based. Microsoft will accept applications for the role until October 21st, 2024. #Cybersecurity #LLM #AIAgents #GenAI #GenerativeAI #AI #KnowledgeGraphs #hiring
Open Threat Research retweeted
My Gradient Boosted Regression Tree model won the most accurate prediction award π€© in the 2024 international #CherryBlossom prediction competition π Yay!!
Notebook: bit.ly/GBRT_notebook
Description: bit.ly/GBRT_description
#OpenToWork @GeorgeMasonU @OTR_Community
ALT Link: https://competition.statistics.gmu.edu/winners-of-the-2024-international-cherry-blossom-prediction-competition/
Open Threat Research retweeted
π Lately, Iβve been fascinated by LLM-based AI Agents, so I built my own from scratch to learn more about them π€β¨
blog.openthreatresearch.com/β¦
I hope it helps and inspires you to build your own π οΈ! #GenAI #LLM #AI @OTR_Community β€οΈ
RT @Cyb3rWard0g: π’ Thrilled to be teaching with @fr0gger_ at @BlackHatEvents #BHUSA π
πBoost your career by learning how to apply Python &β¦
Open Threat Research retweeted
π Last conversation of the day @jupyterthon ! Join us now in a small panel to discuss use-cases of @ProjectJupyter notebooks and Generative AI applied to Security π₯ @clearbluejar @ianhellen @PeteABryan @ashwinpatil @lmeyerov
π΄ Live Now: aka.ms/JupyterthonLiveDay2
