Show HN: MatrAIx – simulate users before reality (Survey/Chat/Web/App)

Researchers from Harvard and MIT released MatrAIx, a population-scale, persona-driven infrastructure for evaluating AI systems and interactive products, featuring a shared schema of 1,290 categorical dimensions and a deterministic, quality-filtered coreset of one million personas available on Hugging Face. The system instantiates sampled persona records as LLM agents across Survey, AI Chatbot, Web, and App environments, with a technical report on arXiv (2608.04205) and an open-sourced Playground and task library on GitHub. The project aims to simulate heterogeneous users for reproducible testing, but the authors emphasize it is not a replacement for evidence from real people.

Simulate before reality. Population-scale, persona-driven infrastructure for evaluating AI systems and interactive products with heterogeneous simulated users. MatrAIx is a population-scale, persona-driven infrastructure for evaluating AI systems and interactive products with heterogeneous simulated users. Instead of testing against a generic or interchangeable user, MatrAIx instantiates sampled persona records as LLM agents and runs them through reproducible tasks across four environments — Survey , AI Chatbot , Web , and App native desktop and mobile, including macOS and iOS . At its foundation is a shared schema of 1,290 categorical dimensions covering background, psychology, capability, and behavior. Personas combine dependency-aware synthetic generation with evidence-aware human grounding; a deterministic, quality-filtered coreset of one million personas is released for research on Hugging Face https://huggingface.co/datasets/MatrAIx2026/MatrAIx Persona 1M Public Release . Shared telemetry, task-owned verification, and reporting connect individual responses and trajectories to subgroup- and population-level findings. The name nods to The Matrix : a simulated world useful for exploration, stress testing, and hypothesis generation, not a replacement for evidence from real people . 2026-08-11 Academic commentary:by Mayank Kejriwal Can we simulate the world? . AI Scientist 2026-08-10 Featured as an X Trending Story https://x.com/i/trending/2086626337561911419 : Harvard and MIT Unveil MatrAIx with 8.3 Billion Virtual Personas . Also covered across tech media, including 36Kr https://eu.36kr.com/en/p/3932853833759876 , Numerama https://www.numerama.com/tech/2308727-ces-chercheurs-ont-cree-83-milliards-dhumains-virtuels-pour-tester-des-produits-a-notre-place.html , Infobae https://www.infobae.com/tecno/2026/08/10/asi-prueba-la-ia-un-mundo-con-8300-millones-de-personas-digitales-matraix-es-el-metaverso/ , AI타임스 https://www.aitimes.com/news/articleView.html?idxno=213824 , CryptoBriefing https://cryptobriefing.com/matraix-simulation-harvard-mit-ai-personas/ , and Startup Fortune https://startupfortune.com/harvard-and-mit-built-an-ai-model-of-83-billion-people-to-test-products-on/ , among others. Also discussed by Cisco VP & CTO Gianpaolo Barozzi https://lnkd.in/p/gE9cV2nw . 2026-08-04 Technical report on arXiv: MatrAIx: Simulating the World with 8.3 Billion Persona Agents https://arxiv.org/abs/2608.04205 2608.04205 . Also featured on Hugging Face Daily Papers https://huggingface.co/papers/date/2026-08-10 paper page https://huggingface.co/papers/2608.04205 . 2026-08-01 Released Persona 1M https://huggingface.co/datasets/MatrAIx2026/MatrAIx Persona 1M Public Release on Hugging Face ~1M quality-filtered personas . 2026-07-31 Open-sourced the Playground and task library: MatrAIx-Persona-8B https://github.com/MatrAIx-ai/MatrAIx-Persona-8B . 2026-07-29 Position note: From Personas to Simulated Users https://matraix.ai/research/survey-from-personas-to-simulated-users.html . Docker https://docs.docker.com/get-docker/ — needed for Web and OS-app tasks uv https://docs.astral.sh/uv/ and Python 3.12- Node.js 20+ Playground / viewer frontends only - Model API keys for real persona runs — see agents.md /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/environment/agents.md the install checks below do not need a key Windows users: run everything inside WSL2 — open PowerShell, run wsl --install installs Ubuntu , then clone this repoinside the WSL filesystem e.g. ~/MatrAIx , not /mnt/c/… , which is much slower and enableWSL integrationin Docker Desktop → Settings → Resources. Every command in this README then works exactly as written. Native PowerShell/cmd is not supported the task verifiers require bash . git clone <repo-url && cd MatrAIx uv venv --python 3.12 uv pip install -e . uv pip install pytest pytest-asyncio httpx uv pip install -e packages/playground uv pip install -e packages/harbor-langsmith uv pip install -e packages/rewardkit Run jobs with uv run matraix run … . After install, use the smoke tests smoke-tests below to confirm Survey, Chat, Web, and OS-app are ready no API key . Summarize a finished job with . Advanced runtime tools stay under uv run matraix results <job uv run harbor … .Set a model API key before real GUI or CLI runs smoke checks do not need one : export ANTHROPIC API KEY="sk-ant-..." anthropic/claude- models export OPENAI API KEY="sk-..." openai/gpt- models See agents.md /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/environment/agents.md for the full key matrix. Playground can also load keys from application/playground/.env.local . The in-repo matraix-persona-dev-sample ~200 is for smoke only. For real cohorts and Playground sampling, import the public 1M coreset: huggingface-cli download MatrAIx2026/MatrAIx Persona 1M Public Release \ --repo-type dataset \ --local-dir persona/datasets/matraix-persona-1m/release Playground: Dataset → matraix-persona-1m . CLI: --dataset persona/datasets/matraix-persona-1m . Details: Handbook § Persona 1M /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/README.md 3-persona-1m-recommended . Two quick checks after install — no API key. Together they cover the default path for all four task types Survey, Chat, Web, OS-app : | Check | Confirms you can run | Command | |---|---|---| Without Docker | Survey and Chat | uv run matraix smoke application/tasks/example-survey product-feedback | With Docker | Web and OS-app | uv run matraix run -c configs/jobs/example-job-recipe/harbor-smoke-local.yaml | The first finishes in seconds and should print Smoke: ok . The second builds a small local image on first run a few minutes , then writes under jobs/harbor-smoke-local/ . Step-by-step: quickstart §3 /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/quickstart.md 3-smoke-tests-two-lanes . Playground picks tasks, samples personas, and launches the same Matraix Playground jobs as CLI auto mode. Start API + frontend two terminals : Terminal A — API VENV=.venv bash application/playground/backend/run dev.sh Terminal B — frontend cd application/playground/frontend && npm ci && npm run dev Open http://localhost:5173 → Playground → pick a persona cohort → pick Survey / Chat / Web / OS app tasks → Lock pipeline → Run eval . Details: Playground §10 /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/quickstart.md 10-playground--play-tasks-visually . Develop — copy a reference task under application/tasks/ , edit task.toml / instruction.md / input/ / verifier, then register it for Playground task-guide.md /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/application/task-guide.md : cp -R application/tasks/example-survey product-feedback \ application/tasks/<your-task-name | Type | Reference task | |---|---| | Survey | application/tasks/example-survey product-feedback | | Chat | application/tasks/example-chat-api support chatbot | | Web | application/tasks/example-web-playwright quote-choice | | OS-app | application/tasks/example-computer-use-linux note-to-csv | Run — generate a Matraix Playground job pins agent + model , then execute it: uv run python application/scripts/generate application job.py \ --task application/tasks/example-survey product-feedback \ --execution-mode auto \ --persona-ids 0042 \ --model-name anthropic/claude-sonnet-4-6 Use the export lines + recipe path the script prints, e.g.: uv run matraix run -c configs/jobs/application-task-job-recipe/example-survey-product-feedback-auto-n1.yaml Batch --sample-size N , filters, and chat / web / os-app examples: docs/quickstart.md /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/docs/quickstart.md . MatrAIx Handbook — guides, persona / application / environment docs. MatrAIx/ ├── persona/ Schema, datasets, synthesis/curation/validation pipelines │ ├── schema/ 1,290-dimension persona schema │ ├── datasets/ Dev sample pool and persona YAMLs │ ├── validation/ Grounding / quality validation suites │ └── scripts/ Persona job & pipeline helpers ├── application/ │ ├── tasks/ Survey · chat · web · os-app task specs │ ├── task-spec/ Shared task contracts │ ├── playground/ Visual runner backend API + frontend │ └── scripts/ generate application job.py and task tooling ├── environment/ │ ├── runtime/ Matraix Playground runtime │ ├── agents/ Persona-conditioned agents │ ├── task-environments/ Docker images / sidecars │ └── adapters/ External adapters e.g. SimpleQA ├── packages/ playground · rewardkit · harbor-langsmith ├── apps/viewer/ Frontend paired with harbor view ├── configs/jobs/ Curated & generated Matraix Playground job recipes ├── docs/ Handbook — persona/ · application/ · environment/ ├── examples/ Minimal example tasks ├── src/matraix/ Python package entrypoints ├── scripts/ Repo-level helpers ├── tests/ Unit / environment tests └── jobs/ Local Matraix Playground run outputs gitignored Large generated datasets stay outside git see the Hugging Face release above . - Join Discord — nickname . Fill the Google Form background, interests, paper authorship / acknowledgements . Full Name - Affiliation - Say hi to us We like to connect you for the shared interest or experience - Participating MatrAIx research community for collaboration or contribution If you use MatrAIx, the Persona 1M dataset, or results from this repository, please cite: @article{li2026matraix, title = {MatrAIx: Simulating the World with 8.3 Billion Persona Agents}, author = {Li, Xiaomin and Hao, Yuexing and Hou, Jianheng and Huang, Jintao and Wen, Qianfeng and Huang, Shirley and Liu, Yifan and Liu, Xiaoyi and Fan, Yilan and Wang, Yijun and others}, year = {2026}, eprint = {2608.04205}, archivePrefix = {arXiv}, primaryClass = {cs.AI}, url = {https://arxiv.org/abs/2608.04205} } Paper: arXiv:2608.04205 https://arxiv.org/abs/2608.04205 · Full authors: GitHub Cite this repository CITATION.cff · Dataset: Persona 1M on Hugging Face https://huggingface.co/datasets/MatrAIx2026/MatrAIx Persona 1M Public Release . MIT — see LICENSE /MatrAIx-ai/MatrAIx-Persona-8B/blob/main/LICENSE .