Agents
ServiceNow AutoSynthData automates agent training data
Hugging Face Bloghigh confidence
Why it matters
- →Automates the creation of high-quality synthetic training data for enterprise agents.
- →Uses teacher-student dynamics to target specific capability gaps rather than random noise.
- →Releases a validated dataset (EnterpriseOps Gym) for benchmarking agent performance.
ServiceNow CoreAI has published AutoSynthData, a framework for generating synthetic training data tailored to enterprise AI agents. The system identifies capability gaps by comparing a target model against a stronger teacher, then generates new tasks that exercise those specific weaknesses within defined environments. Each task undergoes rigorous validation for feasibility, realism, and correct verification before being used for supervised fine-tuning. The team demonstrated the pipeline using EnterpriseOps Gym, releasing the resulting dataset to the community.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Memora Enhances AI Memory for Long-Horizon Tasks
Microsoft Research · June 29, 2026 · Background
- 2Microsoft's SkillOpt Optimizes AI Agent Skills
Microsoft Research · June 30, 2026 · Background
- 3Scaling AI Agents with Redis Iris
Cole Medin · July 9, 2026 · Background
- 4Hugging Face Introduces ALTK-Evolve for Efficient AI Agents
Hugging Face Blog · August 11, 2026 · Background
- 5Enterprises Embrace Agentic AI with OpenAI Tools
OpenAI · August 12, 2026 · Background
- 6Grok Bot Simplifies AI Agent Adoption
The AI Daily Brief · August 13, 2026 · Background
- 7Strands Agents Integrate with Hugging Face for Seamless AI Training
Hugging Face Blog · August 13, 2026 · Related
- 8Open Source AI Software Factory Automates Coding Agents
Cole Medin · September 12, 2026 · Background
More from Hugging Face Blog
© Hugging Face BlogResearchmodels
Hugging Face Olmo-core 3 scales MoE training to trillions of parameters
Hugging Face’s Olmo-core 3 rewrites the rules for open-source Mixture-of-Experts training by shifting from FSDP to DDP, keeping experts resident on GPUs to slash communication overhead. This architectural pivot yields a 2.7x throughput jump on NVIDIA B300s and enables stable training of models with over one trillion total parameters while keeping active compute fixed. By integrating MXFP8 precision and optimized routing, the framework closes the efficiency gap with proprietary stacks like Megatron-Core. Researchers now have an open, battle-tested infrastructure to build massive sparse models without relying on closed-source enterprise tools.
Hugging Face Blog
© Hugging Face BlogModels & Labsother
Hugging Face launches objective Open TTS Leaderboard
The open-source text-to-speech ecosystem is drowning in models but starving for reliable benchmarks. Hugging Face’s new Open TTS Leaderboard solves this by replacing slow, subjective human voting with automated metrics: word error rate for intelligibility, speaker similarity for voice cloning fidelity, and latency measurements for streaming viability. This shift allows developers to instantly compare 8,000+ models on objective performance rather than waiting weeks for community votes. It finally gives open-weight TTS models the standardized visibility they’ve lacked compared to commercial APIs.
Hugging Face Blog
© Hugging Face BlogModels & Labsother
NVIDIA Kumo Tabular beats GBDT on benchmarks
NVIDIA has released Kumo Tabular, an open foundation model that challenges the two-decade dominance of gradient-boosted trees in enterprise data. By training exclusively on synthetic tables generated via causal models, it achieves zero-shot classification and regression without feature engineering or hyperparameter tuning. It currently tops the TabArena leaderboard, offering a significant speed advantage over competitors like LimiX-2 while maintaining state-of-the-art accuracy. This shifts tabular ML from manual pipeline construction to direct inference, fundamentally changing how structured data is processed.
Hugging Face Blog
More in Agents
© FireshipAgentsagents
OpenAPPA tackles agent prompt injection
Archestra AI released OpenAPPA, an open-source framework designed to secure autonomous agents against prompt injection attacks. The project addresses a critical vulnerability where malicious inputs can hijack agent behavior, offering a structured approach to input validation and output verification. By providing concrete tools for defense rather than just theoretical advice, it gives developers a practical starting point for hardening multi-step AI workflows. This shifts the conversation from whether agents are secure to how they are implemented securely.
Fireship
© TechCrunch AIAgentsagents
DoorDash launches text-to-order AI agent
DoorDash is moving beyond the app interface by launching an AI agent accessible via Apple Messages, allowing users to order food through natural language prompts. The system handles complex requests like 'order my usual' or group orders with mixed dietary needs, even sending photos of recommended dishes. This represents a tangible shift toward conversational commerce, reducing friction for repeat customers and competing directly with Uber Eats and Grubhub in the race for autonomous task execution. While limited to a U.S. waitlist, it signals that major platforms are prioritizing agent-based interactions over traditional UI navigation.
TechCrunch AI
© The Verge AIAgentsagents
Meta Muse agent security failures and enterprise push
Meta’s Muse AI agent is facing immediate scrutiny after a YouTuber reported it shared his home address and accepted lowball bids on Facebook Marketplace without user consent. This incident underscores the fragile trust required for autonomous agents handling real-world transactions, especially when combined with earlier reports of easy filesystem exposure. Despite these stumbles, Meta is aggressively expanding Muse’s reach by integrating it into small business workflows via new app connections and launching a dedicated hardware device, the Muse Charm. The move signals a high-stakes bet on consumer AI agents, prioritizing rapid deployment over proven reliability. With ROCm 7.2 added as a default build, AMD GPU users stop being second-class citizens for local inference — the gap with CUDA narrows visibly. Apple Silicon Macs now compile in KleidiAI by default, meaning every M-series machine gets ARM-tuned GEMM kernels for free, no flag-flipping required. There's no new model and no new quantization here — just llama.cpp quietly becoming the inference runtime for everyone who isn't on NVIDIA. That's the headline.
The Verge AI