The Real Problem With Agent Memory Isn't Storage. It's the Contract.
A sharper Medium-style essay arguing that the real differentiator in agent memory is the harness-enforced contract for what gets remembered, recalled, and governed.
Bharat Khanna's Blog
Guardrails as Runtime Policy in Multi-Agent Systems
A systems-oriented look at why guardrails belong inside orchestration, with benchmark results from agentic-runtime across support triage and research-and-retrieval workloads including adversarial cases.
Agentic AI and Data Products: A Framework for Real-Time Autonomous Decision Systems
A high-level framework for integrating agentic AI with real-time analytics pipelines, covering architecture, governance, trade-offs, risks, and evaluation across finance, IoT, and healthcare.
Skills vs MCP: Designing the Next Generation of AI Agent Architecture
Not all problems an AI agent solves are the same. This piece unpacks the distinction between Skills (knowledge layer) and MCP (execution layer) — and why separating them leads to more efficient, maintainable agent systems.
Deploying LangGraph Agents to Production: A Practical Infrastructure Guide
Everything you need to go from a working LangGraph agent to a production deployment: containerization, async serving with FastAPI, observability, circuit breakers, and zero-downtime deploys.
Continuously Improving Agent Quality Using Evaluators Across Single-Turn, Trajectory, and Multi-Turn Interactions
A production-grade guide to agent evaluation: single-turn unit evals, trajectory scoring, multi-turn simulation, harness engineering, and LangSmith-driven continuous improvement loops.