AI / Cloud / Systems
Hiran Patel (HP)
Software architect with 20+ years building systems that scale — distributed infrastructure, DevSecOps, and agentic AI. I write clearly about hard problems.
10 posts
Deterministic Agent Workflows Beat Context-Rot Coding
How code-as-orchestrator, schema outputs, and budget-aware loops outperform ad hoc long-context agent sessions.
I Trusted the Workflow Until It Broke Trust
How a workflow trust failure reshaped my architecture: deterministic orchestration, isolated agents, and portable control.
Tokens Are the Next Spice
Token pricing is repricing AI workloads fast, and teams that model per-token cost without loop depth are underestimating spend.
HIEF: The Memory Layer AI Coding Agents Need
A local-first MCP sidecar that adds durable memory, coordination, and regression-aware quality checks to coding-agent workflows.
Third-Party Libraries: Where They Pay Off and Where They Hurt
How to decide when external dependencies accelerate delivery versus create long-term maintenance and security drag.
Topics All topics →
agileai-agentsclicloudcloud-costscloudadoptioncloudcomputingdeveloper-toolsdiscoverydockerengineeringhtmlinfrastructurekuberneteslibrariesllmopsmarkdownmcpociopen-sourceoraclephysicsproductivityrustsecuritysoftware-architecturesoftware-qualitysoftwaredevelopmentsuperconductorssupply-chain-securityworkflow-automationwriting