Armonti D.

About Armonti D.

I build deterministic control planes for AI systems.My work focuses on governance, auditability, and reproducible oversight in environments where AI decisions must be explainable, replayable, and defensible under scrutiny.I design systems that operate at the boundary of learned models. They evaluate proposals, enforce formal constraints, and generate cryptographically verifiable audit trails without modifying model internals. Determinism, correctness, and failure transparency are treated as first-order requirements, not afterthoughts.My background is in Rust and Python systems engineering, with emphasis on formal invariants, cryptographic lineage, and bit-exact replay over stochastic model behavior. I build assuming hostile inputs, regulatory review, and the need to re-examine decisions long after deployment.I am interested in safety and evaluation infrastructure, deterministic oversight, and how governance mechanisms hold up as AI system capabilities increase.Platform development and validation occurred within Microsoft AI Cloud Partner and ISV Success environments. Translational exploration was conducted in collaboration with the WMED Innovation Center.

Experience

1. Applied Research Builder (Haltra Prototyping)

   Builders Backers

   Aug 2025 — Present

   Used as an applied experimentation environment to prototype and validate components of Haltra’s governance and decision economics stack. Tested enforcement logic, cost modeling assumptions, and decision thresholds against real-world constraints. Refined system boundaries, failure modes, and integration surfaces prior to production hardening at Only SG Solutions.

Education

• Siena Heights University

  BBA, Business Administration and Management

• Eastern University

  M.S. Data Science , Machine Learning (coursework completed)

• University of Michigan-Flint

  M.B.A., Computer Science & Finance (coursework completed)