hsnlabs / Blog
All Posts Agentic Bootcamp™ Advisory About Hugo
hsnlabs
EN · PT
Explore Agentic Bootcamp™
Hugo S. Nascimento
Technical Leadership
Share it with Hugo, CPTO
Start a Conversation

Recent Posts

  • LangSmith and Production Evals: Measuring Agents Before Failures Hit Users
    1 day ago
  • Loop Engineering: Controlling Production AI Agents Without Runaway Costs
    2 days ago

Categories

  • Why Agents Fail
  • Agentic AI
  • Future of Work

Tags

why-agents-fail case-studies production guardrails ontology langsmith evaluation deep-agents observability loop-engineering langgraph context-engineering agent-memory rag ai-agents langchain enterprise architecture comparison agent-harness future-of-work economics c-suite databricks data-contracts snowflake canvas enterprise-agents methodology crm model-context-protocol saas bpo roi ebitda mcp python database sql vector-database ai-failures knowledge-graph data-engineering operational-ontology deterministic-ai margins consulting sprint artificial-intelligence philosophy hr collections hr-tech insurance arbitrage erp-support helpdesk rpa automation poc latam-airlines llm-judge audit api-loops erp legacy-core mainframe schema-drift state-machines legacy-it reliability hsnlabs manifesto
All Posts

Tag: langsmith

Hugo S. Nascimento
Have a project in mind?
Share it with Hugo, CPTO
LangChain Production ↗ Snowflake Partner ↗ Databricks Partner ↗
Start a Conversation
Back to All Posts Explore Agentic Bootcamp™ →
LangSmith and Production Evals: Measuring Agents Before Failures Hit Users

LangSmith and Production Evals: Measuring Agents Before Failures Hit Users

2026-11-15 · Why Agents Fail

How LangSmith observability, targeted trajectory evals, and efficiency ratios catch Deep Agents failures before they hit enterprise users.

Loop Engineering: Controlling Production AI Agents Without Runaway Costs

Loop Engineering: Controlling Production AI Agents Without Runaway Costs

2026-11-14 · Why Agents Fail

How to stack the four loops of loop engineering with LangGraph and LangSmith to prevent runaway token burn, verify outputs, and compound reliability in production.

hsnlabs

hsnlabs is your enterprise AI transformation partner, operationalizing autonomous agents over legacy systems.

AI Architecture Boutique
São Paulo, Brazil · Rua Pais Leme, 215, Pinheiros
hugo@hsnlabs.ai

RSS Feed 2.0 llms.txt for LLMs & Agents
Why Agents Fail
  • What Is an Ontology for AI Agents
  • The Enterprise PoC Graveyard
  • Integration Drift in Production
  • Why RAG Breaks on Financial ERPs
  • LLM-as-a-Judge Fallacy in Banking
Agentic AI
  • Building Enterprise Ontologies from Scratch
  • Perimeter Isolation and MCP
  • Legacy Core as Agentic Engine
  • LATAM Airlines: 3% Margin Engine
  • State Machines vs API Loops
  • Databricks Lakehouse & LangGraph
  • The Agentic Canvas™ ↗
Future of Work
  • BPO Replacement Matrix
  • Extinction of HR Outsourcing
  • Collapse of Legacy RPA
  • C-Suite Margin Playbook
  • Cost of Unbounded AI in Legacy IT
  • Why I Built hsnlabs
  • Newsletter
    [Agentes] x [a Gente]
    [Agentes] x [a Gente]
    1,600+ subscribers
    LinkedIn
    Beyond the hype: how Agentic AI multiplies and leverages human labor capacity.
2026 hsnlabs. Agentic AI Adoption.
GitHub LinkedIn X RSS Feed JSON Feed Sitemap The Agentic Canvas™ Bootcamp Advisory LangChain Snowflake Databricks Braintrust Privacy Policy Compliance