DailyUpdated 2026-11-07 12:20 UTC
Every area we track.
Frontier watches these areas of AI engineering every day, scores what happens in each, and explains what matters.
- Research — Papers and results that change what is possible23 articles · 40 signals
- Agents & orchestration — Agent architectures, multi-agent orchestration, planning loops, tool use, harnesses22 articles · 13 signals
- Evals & reliability — Evaluation, benchmarks, observability, guardrails, reliability of agentic systems15 articles · 18 signals
- Models & training — New frontier and open models, architectures, training, fine-tuning, reasoning9 articles · 17 signals
- Infra & inference — Serving, inference optimisation, GPUs, routing, cost and latency engineering6 articles · 7 signals
- Context & memory — Context engineering, retrieval, long-horizon memory, knowledge for agents5 articles · 7 signals
- Safety & policy — Safety, alignment, regulation, security of AI systems4 articles · 4 signals
- Agentic coding — AI coding agents, spec-driven development, repo intelligence, verification of AI-written code2 articles · 8 signals
- Protocols & interop — MCP, agent protocols, tool standards, interoperability between systems1 articles · 0 signals
- Products & usage — How teams actually ship and use AI products, workflows, adoption0 articles · 6 signals