Build a Convex-Backed Voice Agent with a Reactive Database
Convex's reactive queries auto-push every transcript update to every subscribed client. Real working code for actions, mutations, and OpenAI Realtime streaming.
11 min read
Agentic AI & LLMs
Agentic AI & LLMs archive page 280 of 631
Agentic AI & LLMs
Agentic AI, LLM engineering, and the models behind modern automation — multi-agent systems, LLM evaluation and comparisons, RAG, fine-tuning, AI infrastructure, security, and production AI engineering.
From the blog
Latest articles
CloudEvents 1.0 for AI Agents: One Envelope Across Kafka, NATS, SQS, and Pub/Sub
CloudEvents 1.0 is the CNCF spec for interoperable events. Adopt it as your envelope and your AI agent events travel unchanged across Kafka, NATS, SQS, EventGrid, and Pub/Sub — with SDKs in 9 languages.
10 min read
SBOM for AI Voice Stacks: HIPAA, CISA Guidance, and 2026 Supply-Chain Reality
CISA's 2025 SBOM guidance brings AI and SaaS into scope. Here is the SBOM architecture for a HIPAA-aligned AI voice platform — first-party services, model artifacts, and sub-processor declarations.
10 min read
WebRTC + AI Coach for Live Fitness Classes in 2026: Form Correction at 30 FPS
Live fitness classes in 2026 stream over WebRTC with an AI coach that watches each rider's form via their phone camera and corrects in real time. Here is the production stack with MediaPipe and on-device pose.
10 min
Google AI Principles 2026 — A New CCL on Harmful Manipulation and What It Means
Google's 2026 Responsible AI Progress Report (February 18, 2026) added a new Critical Capability Level focused on harmful manipulation. For voice AI builders, that single change reshapes red-teaming priorities for the year.
10 min read
Twilio Lookup v2 for AI Caller Validation: SIM Swap, Fraud, Identity Match (2026)
Lookup v2 returns line type, SIM swap, call forwarding, identity match, reassigned number, and SMS pumping risk in one call. We use it as the first 200 ms of every CallSphere voice AI session.
10 min read
Where Claude Opus and Claude Code Are Heading Next
How agentic coding with Claude Opus and Claude Code is evolving — longer autonomy, agent fleets, richer MCP — and how to prepare now.
8 min read
How to Measure Success With Claude Opus in Claude Code
The metrics that prove Claude Opus in Claude Code is working — cycle time, rework rate, eval pass rate — and the vanity numbers to ignore.
9 min read
Claude Opus in Claude Code: A Real End-to-End Build
A realistic walkthrough of shipping a feature with Claude Opus in Claude Code: from vague ticket to tested, reviewed, deployed code.
8 min read