Trusted AI answers across disconnected systems.
ArcellAI is an interactive AI data scientist and engineer that connects fragmented mission, engineering, and operational data to produce evidence-linked answers, recommendations, and actions.
Scaffolded by a context-graph-centered agent harness built on a federated unified data architecture.
ArcellAI is a forward-deployed engineering research lab developing an interactive AI data scientist and engineer for mission-critical defense and engineering systems.
We are building the provenance-preserving context and agent infrastructure required to move AI from isolated models into trusted operational workflows.
System Architecture
An active intelligence layer, not another silo.
ArcellAI connects existing systems, documents, models, telemetry, and workflow history through a continuously updated context layer, allowing its interactive agent to reason across authorized evidence without replacing the underlying systems.
ArcellAI Agent & Context Harness
Federated Data & Context
Continuous provenance-graph generation and interactive agent coordination context.
Core Platform Capabilities
Our framework is built to address the high-stakes requirements of mission-critical enterprises where standard AI integration models break down.
Interactive AI Agent
An AI data scientist and engineer that investigates questions, uses approved tools, and produces traceable outputs.
Federated Data Integration
Connects structured data, documents, models, telemetry, and operational systems without forcing a replacement migration.
Mission Context Graph
Resolves relationships among evidence, entities, systems, workflows, decisions, and time.
Provenance and Traceability
Links answers, recommendations, and actions to their underlying sources and transformation history.
Human-Governed Execution
Allows authorized users to inspect, approve, redirect, or reject consequential agent actions.
Model-Agnostic Orchestration
Supports appropriate models and deterministic tools within a governed agent workflow.
Context infrastructure for mission-critical decisions.
ArcellAI helps defense operators, analysts, and engineers work across disconnected data and systems while preserving the evidence, uncertainty, and human decisions behind every output.
Logistics and Supply Networks
Connect fragmented supplier, asset, maintenance, and operational evidence to support resilient planning and dependency analysis.
Mission and Engineering Systems
Trace relationships across technical documents, enterprise systems, engineering models, requirements, and operational events.
Physical AI and Autonomous Systems
Extend governed context and provenance into robotics, sensing, maintenance, simulation, and other physical-system workflows.
Foundry and Partners
Alex Velez Arce
Founder, ArcellAI
Alex is an MIT computer science graduate who developed the core technology powering ArcellAI's architecture as part of his biomedical AI research at Harvard Medical School (sponsored by the Chan-Zuckerberg Initiative). His background includes leadership of Therapeutic Data Commons (TDC) and spinning out PyTDC—an API-first AI and data platform for deploying AI on continuously updated heterogeneous data sources built on rearchitecting TDC—and published research at both NeurIPS and ICML. He won a conference grant for PyTDC's poster presentation at ICML, where ArcellAI was the lead institution on the publication.
Alex has extensive experience building production data infrastructure, ML evaluation frameworks, ML deployment systems, and source-level provenance tracking, acquired at Pinterest, Cruise Automation, and several startups.
Capital & Strategic Support
Capital Backing
ArcellAI is backed by Year Zero Ventures, supporting frontier-technology founders building defensible commercial systems.
Request a Briefing
Tell us about the disconnected data, engineering, or operational systems behind your mission workflow.