{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"ADAPT Insider","title":"Production readiness determines whether AI pilots scale, says Harris Farm Markets’ Head of Infra","html":"","width":"100%","height":180,"duration":732,"description":"AI pilots can create false confidence.\nControlled data, prompts and supervision often make early results look stronger than the systems that eventually reach production.\nLeonard Rawat, Head of Infrastructure, Security and Operations at Harris Farm Markets, explains why infrastructure choices must start with the use case.\nHe examines what changes as token demand grows, why guardrails belong in the architecture and how internal capability turns experimentation into practical value.\nKey takeaways:\nThe use case must determine the infrastructure. Organisations should define the business problem before selecting a platform, model or deployment approach.\nProduction readiness starts during the pilot. Security guardrails, cost controls and token optimisation must be designed before workloads scale.\nBusiness context has to be built internally. Training and sandboxed experimentation help employees turn operational knowledge into useful AI solutions.\n\nProduction exposes the weaknesses pilots conceal\n\nMany AI projects perform well in pilots because the data is clean, prompts are controlled and users receive close supervision.\nThose conditions disappear when systems enter day-to-day operations.\nLeonard says organisations should plan for scale from the outset.\nToken consumption can rise quickly as experimentation develops into agentic workflows or automation, increasing pressure on the platform and its operating cost.\nThe AI platform must align with data location, expected demand and the original business case.\nInfrastructure choices determine whether a promising pilot can deliver reliable value in production.\nGuardrails have to be part of the architecture\n\nAI infrastructure extends beyond keeping systems available.\nIt must account for cloud spend, token usage, security exposure and the wider ecosystem surrounding each model.\nLeonard argues that clear controls are needed at the infrastructure layer before deployment expands.\nSandboxed environments give teams room to test ideas...","thumbnail_url":"https://img.transistorcdn.com/t0Jtrt9eZN4kwrPnSzLt-aLKB9f-tVVg1JGt9pql9t4/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS84ODM3/NGVhYmU2ZjljNWIx/NTliY2NlMTAzZDRl/NGNmOS5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}