Product & Strategy · Enterprise Agentic AI

AI is a product problem, not a technology problem.

Sixteen years at Huawei, MBC Group, and a NASDAQ listed company taught me that. Now I architect enterprise agentic AI for defense, energy, and capital markets, where getting it wrong has real consequences.

See the Work Research at MIT CSAIL

MIT CSAIL Research Affiliate · Kellis Lab · Coauthored an MIT AI study with 271 experts · I still write code most weeks.

Minas Megalokonomos
Dubai · Athens · Boston
Research Affiliate, MIT CSAIL & Kellis Lab
0+Years Executive Leadership
0+Countries Deployed
0Concurrent Ventures Led
$0M+Enterprise BU Scale
01 / Work

Selected Ventures

All launched within the last year. Defense AI, energy, R&D acceleration, and financial infrastructure. I lead each one from first architecture decision through go-to-market.

  1. Agentic AI & Autonomous SystemsChief Product & Strategy Officer

    Multi-Domain Autonomous Detection

    Edge deployable computer vision, the perception layer for agentic AI and autonomous defense systems. The engine fuses up to six sensor modalities, from electro optical and thermal to radar and RF, into real time detection, tracking, and guidance. It runs air gapped on the platform itself with inference under fifty milliseconds, across counter drone systems, ground vehicles, naval patrol, and critical infrastructure. Perception is where autonomy succeeds or fails. I figure out what operators actually need in the field and turn that into working software.

    Computer VisionSensor FusionEdge AIDefense TechAutonomous Systems
    MilestoneDeployed in operational environments we can't talk about here
  2. R&D AccelerationHead of Product

    The Autonomous Research Foundry

    Materials discovery is still painfully slow. Whether it's batteries, polymers, coatings, or advanced composites, the traditional lab cycle takes years. I designed the platform that changes this: Large Quantitative Models (LQMs) that design experiments, paired with autonomous labs that run them around the clock with no human bottleneck. What used to take a decade now takes months. The approach is industry agnostic, and that's the point.

    Deep TechLQMAutonomous LabsMaterials Science
    ImpactAccelerating R&D by 10X across industries
  3. Energy ArbitrageProduct Lead

    Grid Intelligence Platform

    A real-time energy arbitrage engine for solar, wind, and battery storage that trades autonomously on the grid. What gives it an edge: hybrid Alpha Models that combine physics simulation with LLMs, plus live competitor modeling. No human in the loop.

    Energy TechAlgorithmic TradingGrid Optimization
    StrategyAutonomous real-time arbitrage
  4. Capital MarketsChief Product Officer

    IPO Automation Suite

    The traditional IPO book building process, from allocation and pricing to investor syndication, rebuilt as a single secure platform. I designed the architecture and lead ongoing development. Every piece of it has to hold up to institutional scrutiny, and it does.

    FintechCapital MarketsEnterprise SaaS
    CapacitySupporting live listings
02 / History

Before This

Sixteen years of shipping products and building businesses across four continents.

HuaweiLed product strategy within a $100M+ consumer business unit across 20+ countries↗ MBC GroupBuilt the mobile gaming division for the largest TV broadcaster in the Middle East↗ wappierScaled an AI pricing engine for mobile publishers↗ VeltiPart of the team behind Velti's NASDAQ IPO and FPO↗ Boston UniversityBA in Economics & International Finance↗
03 / Research

Bridging academia & industry

For the past three years, I've been a Research Affiliate at MIT CSAIL (Kellis Lab), working on Project Mantis, a platform that maps meaning across data no single model can handle alone: documents, conversations, molecules, entire markets. The same engine that lets researchers search a space of 121 million chemicals for new drug candidates is also organizing classrooms, codebases, and businesses. My focus is the path out of the lab, turning the research into products for the people who need them, in government, healthcare, finance, and beyond. Increasingly that means working with governments directly, helping nations adopt these technologies to accelerate research, sharpen decision making, and grow their own innovation ecosystems. I also mentor students in the lab's AI Rising Scholars community, more than 600 young researchers from around the world.

Selected Publication · 2026
Abstract ↗ PDF ↗
I coauthored this MIT study, one of 272 experts who shaped its findings

Prioritization of Risks from Artificial Intelligence: A Delphi Study of 272 International Experts

Saeri, A. K., Graham, J., Noetel, M., Slattery, P., et al. (2026) · arXiv:2606.04490 [cs.CY]

Three rounds, 24 AI risks ranked by probability, severity, and responsibility for mitigation. One finding worries me more than the rest: the people with the power to make AI safer aren't the ones who'll pay when it goes wrong. The rest of us will. I build this kind of AI, and I treat that as a job description.