I help oil, gas and industrial companies turn field friction into AI-enabled systems for maintenance, supply chain, infrastructure and executive control.
18+ years building technology infrastructure, networks, electrical systems, automation and real operating companies across Argentina, Chile and Italy.
An AI Operations Architect designs the systems that make AI usable in real industrial conditions — not demos. The role sits between IT/OT infrastructure, field execution, supply chain and executive control, turning fragmented operational data into agentic workflows, automated work orders, inventory intelligence and live management dashboards. This page is for operators in oil and gas, energy and industrial services who need digital transformation that survives the field.
Reference: IEA · Energy and AI (2024)The energy sector is moving from disconnected pilots to agentic workflows, digital operations and measurable automation. The winners will not be the companies with the most tools. They will be the companies that redesign field execution, maintenance, procurement, infrastructure and management visibility around reliable data flows.
Field execution is still too manual
Maintenance data is usually fragmented
Supply chain needs earlier risk signals
IT/OT infrastructure defines whether AI can work
Executives need live operational visibility, not late reports
Scroll to move horizontally through five operating zones.
Work orders, inspections, photos and approvals are scattered across WhatsApp, email and spreadsheets.
Mobile workflows, checklists, evidence, signatures, escalation and structured closeout.
Reactive maintenance, weak asset history and slow spare-parts visibility.
AI-assisted work orders, preventive plans, asset timeline, priority suggestions and failure-pattern detection.
Purchasing, inventory and supplier decisions depend too much on memory, manual search and late information.
Inventory intelligence, procurement agents, supplier comparison, contract search and critical-stock alerts.
Remote sites often have unstable networks, unclear monitoring and fragmented device ownership.
Site health monitoring, connectivity redundancy, segmentation, credential control, asset inventory and alerts.
Management sees operational problems after they already became expensive.
Live dashboards for backlog, downtime, risk, cost, stock, compliance and readiness.
Most AI profiles do not understand the field. Most field profiles do not design AI-enabled systems. Jhonatan operates in the middle: infrastructure, electricity, networks, operations, procurement, company building and automation.
Built IT outsourcing before it was common in his market
Built technology retail and services in remote Patagonia
Built infrastructure and connectivity operations across difficult logistics
Built DITAP as enterprise technology infrastructure and automation
Builds SaaS, dashboards, workflows and AI agents from operational problems
Map friction, stakeholders, systems, current processes, data and risks.
Select one pilot. Define workflow, users, data, KPIs, controls and success conditions.
Implement one workflow: work orders, inventory radar, supplier agent, field reporting or site health monitor.
Run with real users, capture exceptions, measure adoption and fix what breaks.
Executive pack with measured value, cost, risk, rollout plan and scale/no-scale decision.
IT outsourcing, networks, servers, support.
Electrical infrastructure and field execution.
Infrastructure and connectivity in Patagonia.
1,200+ SKUs, procurement, technical advisory, operations.
Enterprise infrastructure, managed services, security, automation.
Connectivity intelligence platform.
No. The first step is a controlled diagnostic and one measurable pilot.
No. The first phase scores data readiness and chooses a pilot that can work with available data.
No. It is about reducing manual friction, improving visibility and giving field and management teams better systems.
Only if infrastructure is treated as part of the problem. That is why IT/OT, connectivity and monitoring are included.
Identifying whether there is one operational workflow worth diagnosing in detail.
No generic pitch. We identify one operational bottleneck and decide if a 10-day friction scan makes sense.