Buenos Aires  ·  New York  ·  2026

Operational AI for
enterprise systems.
From  pilot to policy. data to decision. POC to production.

Sergio Mastrogiovanni is a scholar-practitioner who builds AI inside the systems enterprises actually run on — ERP, WMS, control towers — and writes about how to govern it. Founder of Socradata. Postdoctoral Researcher at IAE Business School. Adjunct Professor at NYU.

Sergio Mastrogiovanni — Founder of Socradata, NYU Adjunct Professor, IAE Postdoctoral Researcher
01 · The thesis

Most enterprises generate operational data. Few decide with it. The work is moving organizations from one state to the other — and writing the governance the rest of the market still pretends it does not need.

/01

Build inside the system

AI lives where ERP, WMS, and control towers already run. The integration is the hard part; the model is the easy part. Production-grade decisioning sits inside operational workflows, not in a parallel dashboard.

/02

KPIs before APIs

Measurement design comes before technology selection. Every engagement starts by naming the operational decision that has to change, the KPI that proves it, and the human who owns the override path.

/03

From pilot to policy

Pilots that do not name a route to production are theater. The work is to institutionalize — through governance, MLOps, and interoperability — what a proof-of-concept only suggested.

Research informs delivery. Delivery sharpens teaching. Teaching disciplines research.

Three roles run concurrently. They are not parallel tracks. Each role feeds the other two — that is the practice.

Founder · Principal Consultant

Socradata

Operational AI for enterprise systems. Practice areas span predictive inventory and demand, warehouse and logistics intelligence, supply chain risk and visibility, and AI copilots embedded in ERP and WMS workflows. Buenos Aires & New York. LATAM-anchored, US-extended.

socradata.com · founded 2024Visit Socradata
Postdoctoral Researcher

IAE Business School

Universidad Austral, Buenos Aires. Research stream on data-driven urbanism and smart cities — anchored in CABA and the broader Latin American region. Producing chapter drafts, peer-reviewed submissions, and policy-facing memos.

austral.edu.ar/iae · since 2025IAE faculty page
Adjunct Professor

New York University

NYU School of Professional Studies. Graduate-level teaching at the intersection of intelligent automation, AI, marketing analytics, and digital transformation. Course HCAT1-GC2015. Affiliated researcher, NYU Stern Center for Sustainable Business.

sps.nyu.edu · since 2018NYU faculty page

Four pillars. Six topics. One argument: production over performance.

Keynotes for executive audiences and academic seminars. Available in English and Spanish. Format adapts to context — 45-minute keynote, 90-minute executive workshop, full-day intensive.

PILLAR /01

From POC theater to production AI

Why most enterprise AI pilots stall — and the operating model that gets the next 24 months out of the slide deck and into the warehouse, the planning system, and the audit trail.

Keynote · 45 minEN / ES
PILLAR /02

KPIs before APIs

A measurement-first framework for enterprise AI. Why CDO mandates fail when integration is solved before measurement, and how to invert the sequence without slowing delivery.

Workshop · 90 minEN / ES
PILLAR /03

Operational intelligence for supply chain

Forecasting, anomaly detection, and decision support inside ERP and WMS systems. What works at retail, distribution, and manufacturing scale — what doesn't survive contact with the shop floor.

Keynote · 45 minEN / ES
PILLAR /04

Data-driven urbanism in the Global South

Smart cities without smart governance — what the LATAM evidence shows, where infrastructure-first thinking fails, and how Buenos Aires and the broader region offer a different scaling model.

Academic · 60 minEN / ES
EXECUTIVE /05

Storytelling with data

Executive masterclass on translating analytical work into board-room narrative. Built from the NYU SPS course and refined across multiple LATAM cohorts. Anchored in real enterprise cases.

Masterclass · half-dayEN / ES
EXECUTIVE /06

Intelligent automation & the future of work

NYU SPS course HCAT1-GC2015, adapted for executive audiences. Where automation actually substitutes labor, where it augments it, and the workforce-design questions HR is currently underprepared for.

Masterclass · full-dayEN / ES

An eight-year body of public work. Curated, dated, sourced.

From the NYC Tech Summit at NYU in 2018 to the NYU Future Workforce Global Summit in Buenos Aires in 2026. Conference talks, podcast appearances, op-eds, and academic lectures across English and Spanish, in the United States, Argentina, and the United Arab Emirates.

2026 NYU Future Workforce Global Summit

2025 Academic year · IEOM presentations

2023 Generative AI era

2022 12 talks & columns

2021 ThinkFuture · ITBA · IHRIM

2020 11 sessions · pandemic year

2019 NYC · Abu Dhabi

2018 First public talk

Ongoing Affiliations & bodies of work

Peer-reviewed papers, working drafts, and field notes — built to be cited and used.

Four peer-reviewed papers in 2025 (IEOM / ResearchGate). Two earlier industry articles for IHRIM. Two 2026 works in progress at IAE — one on operational AI institutionalization, one on data-driven urbanism in the Global South.

2026Working paper · IAE

From pilot to policy: institutionalizing operational AI in LATAM enterprises

IAE Business School · in preparation

In progress · IAE Postdoc workstream
2026Chapter draft

Data-driven urbanism in the Global South: governance frictions and the path to scale

Edited volume on smart cities · forthcoming

In progress · IAE Postdoc workstream
2025Peer-reviewedIEOM

AI-Driven Data Governance for Smart Cities

IEOM Society · ResearchGate index 392429464

View on ResearchGate →
2025Peer-reviewedIEOM

Participatory AI Design in Smart Cities — China & the U.S.

IEOM Society · ResearchGate index 395866850

View on ResearchGate →
2025Peer-reviewedIEOM

Navigating Digital Transformation in LATAM Cities

IEOM Society · ResearchGate index 394230019

View on ResearchGate →
2021IndustryIHRIM

AI and the Future of Talent Management

IHRIM · industry article

Read on IHRIM →
2021IndustryIHRIM

AI 101 — A Primer in Artificial Intelligence

IHRIM · industry article

Read on IHRIM →
So what: the research is a feedback loop, not a sideline. Every consulting engagement informs a paper; every paper sharpens the next engagement. The papers are not the deliverable — the framework that travels between them is.

Outcomes, not outputs.

Three representative engagements. Real production AI inside enterprise systems — not pilots, not dashboards, not POC theater. Numbers anonymized; methodology cited on request.

Retail · Distribution

Inventory write-off prevention

$50M
prevented write-offs · 12-month rolling

Predictive layer on top of legacy ERP. Forecast-error decomposition fed back into category-level safety stock and PO timing. Productionized inside the existing planning workflow.

Logistics · Warehouse

Warehouse intelligence platform

35%
operational efficiency gain · pick & labor

Picking optimization and labor forecasting layered on the WMS. Anomaly detection on slot misallocations. Decision support — not autonomous action — with explicit human override.

Manufacturing · Demand

Demand forecasting overhaul

15%
forecast-error reduction · 6-month sustained

Replaced the legacy moving-average baseline with a hybrid statistical and ML stack. KPI defined first; integration second; model third. Survived three planning cycles unchanged.

Selected commentary in regional and US press.

Forbes Argentina

El futuro necesita CEOs audaces (Gemelos digitales)

2022 · columnRead column →
Infobae

La nueva normalidad: una oportunidad para empresas, gobiernos y la comunidad

2020 · op-edRead op-ed →
Ámbito

Industrias 4.0: la recuperación será digital

2020 · op-edRead op-ed →
Correlation One

DS4A · How AI is helping business thrive in unprecedented times

2020 · webinar featureView feature →
ThinkFuture · podcast

Ep. 420 — Intelligent Automation & Sustainability

2021 · podcastListen →

Two doors in.

For an enterprise initiative inside an operational system, start with a 14-day diagnostic. For a conference, executive briefing, or academic convening, reach out for a speaking or advisory engagement.

Door 01 · Enterprise

Diagnostic discovery — 14 days

For executives evaluating an operational AI initiative inside an ERP, WMS, or supply-chain platform. Two-week engagement, fixed scope, fixed price. Output: a productionization roadmap with named KPIs, governance touchpoints, and a path from pilot to policy.

  • Operational decision & KPI definition
  • System & data integration assessment
  • Governance, MLOps, and human-in-the-loop design
  • 30 / 60 / 90-day production roadmap
Book a working session
Door 02 · Speaking · Academic

Speaking, teaching, advisory

Keynotes, executive workshops, academic seminars, and policy convenings — in English or Spanish. Available for in-person events in LATAM and the United States, and virtual sessions globally. Topics span the four speaking pillars and the executive masterclasses.

  • Conference keynote · 45 minutes
  • Executive workshop · 90 minutes to half-day
  • Masterclass · half-day to full-day
  • Academic seminar · 60 minutes
Send a speaking inquiry
So what: direct contact, two clear paths, no funnel. Email is checked daily. Every introduction is read by Sergio personally — not a chief-of-staff, not a form router.