José Macías Silva

Experience

• Leading AI integration at OG&E. Building the AWS-hosted LLM integration layer to connect Claude to operational data across ADMS modules (OMS / SOM / FLISR), CAD, and Resource Availability — API gateway, role-aware context manager, per-query audit logging. Four production-bound LLM capabilities in flight: (a) Cause Code QA — AI validates CAD notes against assigned cause codes; (b) Switching Dashboard AI assist — AI validates SOM switching entries against procedures; (c) Resource Availability natural-language interface — plain-language updates by non-technical users; (d) Nighthawk AI BI — anyone describes analytics needs; Nighthawk generates the dashboard. Also author the labor-cost models and capital-justification artifacts that translate each AI capability into a fundable work order — the bridge between AI engineering and capital approval that operations teams typically can’t produce. Follow-on adds ML predictive insights (ETR, ADMS) and agentic-AI workflows under a four-tier governance framework I authored. Impact targets four pillars: reliability (SAIDI lever), cost (O&M reduction), safety, and regulatory compliance.
• Designed and built the Resource Availability Suite, an in-house enterprise web platform linked to OG&E’s internal systems (Azure AD, ADMS, CAD, SAP HANA), consolidating 5–7 fragmented scheduling sources into one source of truth. Built specifically to OG&E’s operational process — change management is fast and adoption by adjacent departments is straightforward. Architected a four-strategy conflict-resolution algorithm with full audit restoration. Modeled $284K–$310K annual O&M savings · 4,549 hours/year saved; dispatch decisions reduced from 5–7 min to 30 sec (91% faster); 90% reduction in wrong dispatches. External SI quoted $800K for the same scope; delivered internally in 10 weeks for ~$40K (~$760K savings vs. external quote).
• Authored Nighthawk’s 5-year platform roadmap — the strategy for turning fragmented operational tooling into a unified AI-native platform. Phase 1 (Foundation, complete): consolidated 8+ legacy tools under SSO with three-tier RBAC. Phase 2 (in flight): production AI integration. Phase 3 (2028–2030): autonomous agentic workflows. Phase 4 (Enterprise vision): cross-organization platform extension. The pattern transfers to any operations-heavy company running on legacy systems.
• Build complex analytics dashboards that translate operational data into business-critical decisions. Strategic Undergrounding Analytics maps every overhead single-phase lateral on a Google Maps frontend and ranks each as an undergrounding candidate via a composite ROI / vegetation / reliability / infrastructure score (9+ installation-method cost models). Operator Switching Dashboard measures organizational safety through a peer-anchored MAD-based scoring model — pass-through percentage as the leading indicator of switch-order quality before field deployment. Replaces anecdotal capital and safety decisions with data.
• Served as data SME on the ADMS upgrade, validating future-state schema and data flow against five downstream consumers (Oracle, HANA, iDashboard, PowerBI, Nighthawk). Performing this scope internally rather than via contractors generated $49,031 documented cost avoidance; zero-disruption cutover.
• Modernized the Circuit Operations App (SCADA operations + alarm management) from a legacy C# desktop application to a web application on the Nighthawk platform — removing per-workstation install dependency and expanding access from a handful of installed users to anyone on the OG&E intranet.
• Presented “AI at OG&E — Transforming Grid Operations” at the Southeastern Electric Exchange (S.E.E.) DOCC Committee Spring Meeting (April 2026) to peer utility leaders. Presented OG&E’s three production applications built using the seven-step AI-assisted development methodology.

• Pioneered the organization’s first Sales-Inventory-Operations Planning process for a $200M+/yr revenue line across 10 product lines; led Kanban / PFEP implementation reducing inventory holdings while sustaining zero stockouts.

• Designed and built a web-based BI platform serving 9 departments; consolidated four ERP systems (SAP, JDE, Oracle, Quantum) into a single SQL data layer; authored 12-month rolling demand forecasts informing the 10-year strategic plan.

Earlier Aerospace Experience · 2009 – 2011
Pacific Scientific HTL (Program Analyst) · Pacific MultiTech (Program Analyst). Aerospace qualification test procedures, ROM / FFP proposals, sales pipeline forecasting.

Skills

Education

B.S. Aerospace Engineering · California State Polytechnic University, Pomona
Certificate in Management for Technical Professionals · University of California, Irvine