Selected work 5 min read
A coordinated architecture for corporate valuation
NYIC leadership experience connecting predictive assumptions, five valuation methods, evidence retrieval, and investment review.
↗
NYIC · Research & engineering
Expanding what intelligence can do.
Engineering what it takes to make it work.
The next frontier is built.
NYIC advances research in language, world, and multimodal models and engineers production-grade AI systems for enterprises.
NYIC PERSPECTIVES / 01Featured perspective · World models
To reason about the world, a system needs more than a description of it. It needs a model of how it changes.
Explore learned environments, action-conditioned prediction, and the difficult question of when a simulated future can be trusted.
Read the perspective01 / Research
Six connected research areas. One ambition to understand and extend the capabilities of intelligent systems.
Investigating how language models acquire representations, adapt to specialist work, and reason reliably with evidence.
02Studying learned environment representations, action-conditioned dynamics, planning, and uncertainty over time.
03Connecting language, images, audio, video, and documents through grounded representations and traceable evidence.
04Investigating planning, tool use, memory, and coordination in systems that act over extended workflows.
05Developing evidence about model capability, robustness, uncertainty, and behavior within accountable systems.
06Investigating adaptation, compression, serving, and compute allocation under explicit quality and latency requirements.
02 / Applied AI
Production AI is a systems discipline. Models, knowledge, tools, infrastructure, and human judgment have to work together.
Explore our capabilities03 / Selected experience
Selected work 5 min read
NYIC leadership experience connecting predictive assumptions, five valuation methods, evidence retrieval, and investment review.
↗Selected work 5 min read
NYIC leadership experience integrating earnings audio, documents, market information, and specialized research agents.
↗Selected work 5 min read
NYIC leadership experience combining clinical document retrieval, governed analytical queries, and specialized patient-context tools.
↗04 / Industries
The most consequential AI problems sit inside a specific business, a specific workflow, and a specific set of responsibilities.
Our industry perspectives05 / Insights
Insights 6 min read
Where learned dynamics can improve operational simulation, and how to establish a useful validity boundary.
↗Insights 6 min read
An engineering framework for delegated identity, scoped tools, approval boundaries, and accountable execution.
↗Insights 6 min read
Connecting task quality, human effort, adoption, and operating cost to a defensible AI investment case.
↗The NYIC approach
01 / Working together
Define a hypothesis, an experimental approach, and the evidence needed to distinguish progress from a compelling demonstration.
Model behavior · world dynamics · multimodal reasoning
Explore an engagement02 / Working together
Translate a consequential workflow into model, evidence, tool, and infrastructure requirements—with acceptance criteria from the start.
Architecture · implementation · integration
Explore an engagement03 / Working together
Trace a quality, reliability, latency, or cost problem to its cause. Establish a baseline, test the changes, and define a controlled release.
Evaluation · reliability · inference
Explore an engagement