Projects I’ve
worked on.
Two decades architecting intelligent systems for Fortune 500 enterprises. Innovation labs that ship production-scale platforms across finance, healthcare, media, luxury, and enterprise software.
Operating thesis
The next decade of competitive advantage will be won by enterprises that can compress the distance between frontier AI research and shipped, governed, revenue-generating products. That compression is what Nikos and his teams build.
A working record
across two decades.
Shipping AI before it was mainstream — early NLP and predictive modeling through today's generative and agentic platforms.
15+ years deploying production AI/ML to B2B and B2C audiences serving millions of users.
Designed and launched corporate AI labs for major technology companies.
Operating frameworks for rapid experimentation, responsible governance, and translation of research into durable advantage.
Research at the convergence of AI, quantum, cognitive science, economics, and planetary-scale computation.
Bridging theoretical depth with pragmatic implementation across industries and disciplines.
Twenty years in systems engineering. Award-winning author of Mindshare.
Strategic business acumen paired with hands-on technical expertise across product, R&D, and operations.
Three practices,
one operating system.
Frontier ML,
applied.
Advanced AI research and deep-tech expertise across disciplines and verticals — from frontier ML architecture to applied research re-envisioning human–machine modalities.
Design through
augmentation.
Pioneering AI-augmented design across the entire development cycle. Strategic integration of AI into creative and media workflows for luxury brands, A-list studios, and cultural institutions.
Ship, incubate,
invest.
Building new AI products, platforms, and solutions while incubating and investing in high-growth startups. End-to-end commercialization, from POC to production at enterprise scale.
Selected work,
2006 — 2026.
Aion Research Corp.
Plate 04 · Flourish OSChallenge
Institutions navigating the AI transition face questions too technical for traditional strategy firms, too consequential for generic AI tools, and too dynamic for static research reports. Faith-based, fiduciary, and mission-sensitive buyers need governed intelligence — auditable sources, versioned models, explicit uncertainty.
Approach
A hybrid intelligence company built on one evidence graph, model factory, and governance layer — combining analyst-led research, validated data models, simulation-driven decision support, and retained advisory. The first wedge is Flourish OS: a custom research platform for Bountiful Financial and aligned faith-based institutions studying how AI shapes human flourishing.
Decision
A six-layer commercial stack — public research, paid research, institutional data models, simulation tools, standards, and advisory — where every public insight points to a paid product, every product to an institutional model, every model to advisory, and every engagement back into the evidence graph.
Bountiful Financial.
Challenge
Faith-aligned investors need defensible controversy intelligence, faith-enhanced index reporting, and a production data architecture — but the ground state is fragmented vendor pipelines, manual processes, and methodology that lives in spreadsheets. The institutional buyer wants fact sheets, exclusion summaries, and case-level evidence they can put in front of a committee.
Approach
Stand up Bountiful Intelligence — an AI lab built on a unified data spine. One Q2–Q3 program ships four interlocking streams: Unified Data Architecture v4.7 (six canonical tables, daily SFTP and quarterly rebalance pipelines, exclusion diff engine, API layer); Living Case Commercial MVP (case-centric controversy intelligence, evidence hub, watchlists and alerts, methodology center); Index Intelligence Reporting across three calc partners (S&P DJI, MerQube, LIXX) with ISIN-keyed entity resolution; and an Azure production posture for SOC 1 readiness and S&P vendor onboarding.
Decision
Every public output — fact sheet, exclusion summary, case page — traces back to the canonical schema; the methodology knowledge graph is the proprietary moat, exposing pre-rendered reasons rather than raw rules. A bench-to-paid-pilot motion converts beta validation into ARR.
Plate 09 · Faith-aligned intelligenceDirecTV.
Challenge
DirecTV had been losing customers for several years and could not identify the underlying drivers from existing structured data. They engaged Lyrical to mine online customer conversations and surface the real signal.
Approach
We analyzed millions of churn-related conversations to isolate specific interactions where former customers explained why they left. Two-thirds of churn traced to signal loss during weather events disrupting football telecasts.
Decision
A retention plan tightly mapped to those interactions — and a continuing methodology to surface new churn drivers as they emerged.
JetBlue.
Challenge
JetBlue needed to translate fragmented customer signal — across gates, terminals, in-flight experience, baggage, and crew interactions — into a concrete, prioritized brand-experience program above and below the wing.
Approach
AI-driven customer intelligence ranked opportunities across the full journey: gate comms, TSA flow at SFO, signage, AV, seasonal snacks, delay kits, and crew behaviors — alongside spare-parts positioning, scheduling, and baggage operational changes.
Decision
A redesigned Long Beach terminal and new customer-centric executive metrics aligned the organization around a unified experience standard.
End-to-end —
frontier research through
full-stack production.
AI research & engineering
Legacy data science and ML through advanced LLM architecture, multi-modal systems, and generative creation.
Agentic operating systems
Agentic orchestration of business, product creation, research intelligence, workflow automation, predictive modeling.
Market intelligence
Competitive, sentiment, controversy, and perception analytics across an integrated media ecosystem.
AI Media Lab
Proprietary IP and creative production with world-renowned research fellows, AI artists, and practitioners.
Creative & design studio
AI integrated into creative and media workflows — concept to final production — for luxury and cultural institutions.
Ethics, governance & sovereignty
Ethics research, policy development, responsible implementation in partnership with a global think-tank network.
Full-stack applications
From on-prem data center through large-scale multi-cloud (AWS / GCP / Azure) deployments at organization scale.
Strategy & transformation
Global consultancy and strategic deployments that have generated $500M+ in measured business impact.
Venture & investments
AI startup incubator and venture arm. Advisory across early-stage, M&A, and late-stage acquisition.
An unusually broad
cross-section of category leaders.
& Insurance
& Travel
Consumer Brands
& Technology
& Wellness
& Civic
My research explores emergent intelligence at the convergence of AI, quantum computing, cognitive science, economics, and planetary-scale computation — bridging theoretical depth with pragmatic implementation.
A work of mythic science fiction
Ancient stories. Planetary intelligence.
AmericanMythology
A descent into the hidden machinery of power. A search for a future worthy of the people who will inhabit it.
What remains human
when everything can be modeled?
On attention, intelligence,
and the architecture of a republic.
Mindshare — the future of attention, intelligence, and human-machine collaboration.
Used as a reference text by enterprise innovation teams designing next-generation customer experiences.
Universal Data Sovereignty
Presented as part of the Diplomatic Courier delegation — the architecture of a Republic of Intelligences, drawing on Heidegger and Bratton.
Rocket Fuel · Sizmek
Operating experience through major adtech consolidation — a working view of how AI assets are valued, integrated, accelerated, or stranded.
Start a conversation with Nikos
The next AI wave is
enterprise-grade, governed, agentic.
What CTOs take away
- 01A calibrated read on which AI investments will compound — versus which become technical debt within 18 months.
- 02A reference architecture for agentic operating systems that integrates with existing data, security, and governance estates.
- 03A roadmap balancing speed-to-value with responsible governance and regulatory readiness.
- 04Introductions across Lyrical's research, design, and venture network — foundation-model partners and specialist talent.
- 05A candid view on build-vs-buy-vs-partner across the categories the entity competes in.
Discussion topics
- AWhere each entity is on the curve from POC through MVP to production AI.
- BAgentic OS patterns that work inside regulated and brand-sensitive enterprises.
- CGovernance, ethics, and data sovereignty — designing for the next regulatory wave.
- DGenerative AI in creative and media workflows, with luxury and enterprise-grade discipline.
- EJoint research, venture, and commercialization opportunities across the DMC portfolio.