Adam Fourney
@adamfourney
Followers
743
Following
960
Media
57
Statuses
1K
Senior Principal Researcher; Human-AI eXperiences (HAX); Microsoft Research. AI agent wrangler. Building @pyautogen, #MagenticOne, #MagenticUI, and #markitdown
Redmond, WA
Joined October 2009
Microsoft Research has introduced Magentic Marketplace, an open-source simulation environment designed to study agentic markets where LLM-powered agents act on behalf of users or businesses. Unlike past limited experiments, this environment models dynamic, multi-agent
1
6
16
If agents can't collaborate, can they truly participate? Turns out there's more to do before agents can contribute to our connected society. w/ @im_td, @adamfourney, @cervisiarius, @erichorvitz, @ecekamar 👇
We’ve identified a “Collaboration Gap” in today’s top AI models. Testing 32 leading LMs on our novel maze-solving benchmark, we found that models that excel solo can see their performance *collapse* when required to collaborate – even with an identical copy of themselves. A \🧵
0
2
8
Check out our recent work on Magentic Marketplace! w/ @bansalg_ @HuaWenyue31539 @ZacharyHuang12 @adamfourney @ASwearngin77874 @w_epperson T. Payne @jakehofman @BrendanLucier @chisingh @markusmobius A. Nambi A. Yadav @davmicrot @MayaMurad0 M. Vogel @rao2z @erichorvitz and others!
Must-read AI research of the week: ▪️ Supervised Reinforcement Learning (SRL) ▪️ SPICE: Self-Play In Corpus Environments Improves Reasoning ▪️ Reasoning-Aware GRPO using Process Mining ▪️ Magentic Marketplace: An Open-Source Environment for Studying Agentic Markets ▪️ The Era of
0
5
13
AI agents are transforming digital marketplaces, mediating discovery and transactions between consumers and businesses. The new Magentic Marketplace provides an open-source, extensible simulation environment for studying different agentic market designs: https://t.co/UET0smOaJ8
2
16
63
We’ve identified a “Collaboration Gap” in today’s top AI models. Testing 32 leading LMs on our novel maze-solving benchmark, we found that models that excel solo can see their performance *collapse* when required to collaborate – even with an identical copy of themselves. A \🧵
3
25
63
Must-read AI research of the week: ▪️ Supervised Reinforcement Learning (SRL) ▪️ SPICE: Self-Play In Corpus Environments Improves Reasoning ▪️ Reasoning-Aware GRPO using Process Mining ▪️ Magentic Marketplace: An Open-Source Environment for Studying Agentic Markets ▪️ The Era of
6
74
397
🚀 Excited to share our new research from MSR! 🎯 We built "Magentic Marketplace" - an open-source platform simulating what happens when BOTH consumers AND businesses use AI agents for transactions. With ChatGPT shopping for us and AI handling customer service, what does this
0
9
35
For this Halloween, I am going 🎃(m)Agentic!🎃 Enjoyed collaborating with the many impressive folks at @MSFTResearch on their Magentic Marketplace ( https://t.co/kpYViq0GaQ), an open-source testbed and evaluation of multi-agent marketplace that has LLM agents acting on behalf of
🌻 Announcing New Agents + Economics Research from Microsoft! AI agents are starting to shop and buy for us. At the same time, agents are representing and providing customer support on behalf of businesses. We believe that these two sides will soon collide, and... 1/n
3
6
27
🌻 Announcing New Agents + Economics Research from Microsoft! AI agents are starting to shop and buy for us. At the same time, agents are representing and providing customer support on behalf of businesses. We believe that these two sides will soon collide, and... 1/n
2
11
19
Read more about our work in our blog post https://t.co/3LLiSKATy4 and try it today in https://t.co/jvKAqVb2SG or via `pip install magentic-ui`, our benchmark SentinelBench is still under development but will hopefully be out soon, work with @adamfourney @MayaMurad0 @bansalg_
0
3
5
I am extremely excited about this work! It opens a new and useful way of using agents (addressing a new category of "FOMO" tasks). And, it's highly generalizable: monitor prices, paper pubs, GitHub, Discord -- even traffic cameras, all with a simple text prompt
SentinelStep enables AI agents to handle monitoring tasks that run for hours or days, like watching for emails or tracking prices. It works by managing when agents should check and their context, avoiding wasted resources and missed updates. https://t.co/DtW4nfDQtJ
0
3
4
✈️ Planning to attend #CHIIR2026 in Seattle? Getting your visa invitation letter is quick and secure! Just email chiir26-registration@easychair.org and you’ll receive a secure link to generate your letter instantly. 🔗
0
3
4
LAST CALL 🚨 #CHIIR2026 Full Paper Abstracts close TODAY (AoE)! Bring your research on: 🤖 #LLMs & #RAG 🧠 Human factors & cognitive models 💬 Conversational search & interaction 🌍 Data ethics & user experience Submit now ➡️ https://t.co/mrhMCYLTLU
0
2
4
Working on #LLMs, human-AI collaboration or interactive retrieval? CHIIR is your venue --> IR, #HCI, and human sciences! 📅 Only two days left to abstract deadline submission https://t.co/mrhMCYLTLU
#SIGIR
0
3
6
To all AutoGen users, please read the following letter.
0
4
9
Two years after @pyautogen first release, it is now officially graduating into the Microsoft Agent Framework — a convergence of AutoGen and Semantic Kernel into a single, pro-code, production-ready SDK for multi-agent systems.
AI Economics depends on efficient token factories and highly performant agent frameworks that deliver enterprise outcomes! That is why we are excited about Microsoft Agent Framework. You can now build, orchestrate, and scale multi-agent systems in Azure AI Foundry using this
0
4
12
The CHIIR community is growing and we want your voice in Seattle! 🌲 📄 Abstract deadline: Oct 8, 2025 📄 Full Paper deadline: Oct 15, 2025 Be part of the conversation on how humans search, explore & interact with information. 👉 https://t.co/mrhMCYLTLU
#CHIIR2026
0
3
8
Whatever your contribution is, we’ve got a track for you! At CHIIR 2026, you can submit: 📄 Full & Short Papers 📚 Resource & Perspective Papers 💻 Demos 🛠 Tutorials & Workshops 🎓 Doctoral Consortium Details 👉 https://t.co/mrhMCYLTLU
#CHIIR2026 #LLMs #HCI
0
6
7
MCP is an exciting start to future agentic ecosystems but is it enough? The current MCP ecosystem is vulnerable to tool interference. We opensource MCP reviewer as a tool to help.
microsoft.com
As agentic AI ushers in a new era marked by tool expansion, systems are converging, and complexity is rising. Microsoft Research explores the Model Context Protocol (MCP) as a new standard for agent...
1
2
6
🤖Tools empower agents, until they clash. 🛠️ Check out our new blog post 👇 on tool-space interference in today’s #MCP ecosystem, and how our open-source✅MCP Interviewer can help you catch MCP server issues before your agents do! #AIagents #MicrosoftResearch
As agentic AI ushers in a new era marked by tool expansion, systems are converging, and complexity is rising. Microsoft Research explores the Model Context Protocol (MCP) as a new standard for agent collaboration across fragmented tool ecosystems. https://t.co/fOxgng3FkI
1
1
5