MemGPT in Production 2026: Hard Lessons from Real Deployments
A year of MemGPT (now Letta) in production has taught teams hard lessons. The five patterns that stop your memory layer from drifting and burning your budget.
10 min read
Agentic AI & LLMs
Agentic AI & LLMs archive page 298 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
Salesforce MCP Server: Sales Agents on Live CRM Data Done Right
Salesforce's MCP server gives agents safe read/write access to leads, opportunities, and accounts. The architecture and the field-level security gotchas that bite teams.
10 min read
Cloudflare Workers + Durable Objects at 10k Concurrent: Real Cost
We modeled 10,000 concurrent voice agent WebSockets on Cloudflare. With hibernation and the 20:1 message ratio, the bill lands surprisingly low. Here is the line-by-line math.
11 min read
Chat Agent Prompt Versioning and Rollback in Production: 2026 Patterns
Production prompts change constantly and break quietly. Here is how to version, deploy, and roll back chat agent prompts in 2026 — with instant revert and zero redeploy.
10 min read
Synthetic Monitoring for Voice Agents: Checkly, Datadog, and Building Your Own
Real users generate noise. Synthetic checks generate signal. Here's how to run a fake voice call against your agent every minute and catch regressions before customers do.
11 min read
RTP Transcoding Cost for AI Voice in 2026: Why Edge Placement Beats Central GPU
Transcoding RTP to WebSocket is more CPU-intensive than people expect. For AI voice in 2026, where you place the transcode (edge near the carrier vs central near the model) decides your cost-per-minute.
11 min read
Helicone OSS vs Cloud in 2026: When to Self-Host Your AI Gateway
Helicone has processed 2B+ LLM calls and ships both as a managed cloud and as fully self-hostable open source. Here is the actual decision tree for 2026.
10 min read
Replace Rosie AI With a HIPAA-Grade Voice Agent for Healthcare
Rosie's $49–$299 plans target micro-businesses and don't sign BAAs. For dental/clinic use you need HIPAA controls — here's the full build.
13 min read
Pipecat Framework for AI Voice Pipelines in 2026: 100+ Services, Frame-Based Orchestration
Pipecat is the Python-first framework for STT-LLM-TTS pipelines with 100+ AI services as plugins, ultra-low latency, distributed Subagents, and direct Twilio integration. Here is the 2026 build pattern.
11 min read