Computational Trust · Deep Machine Learning · Sovereign AI

Advancing Trustworthy Machine Intelligence for Canada's Deep-Tech Future

I investigate the computational foundations of trust in machine intelligence, developing and evaluating deep learning methods, reliable autonomous systems, and secure AI architectures. My evolving research agenda connects rigorous machine learning with Canada's technological capabilities.

Explore Research AgendaCollaborate

About

I am a computational trust scientist, researcher, and educator working across machine learning, information trust, trustworthy AI, cybersecurity, and human-centred computing.

My earlier research on information verification and computational trust provides the foundation for a new program investigating how trust can be modelled, measured, learned, and used in intelligent decision-making. The aim is to contribute novel methods and reproducible evidence—not simply governance principles—for reliable machine intelligence.

I earned a Ph.D. in Computer Science at Ontario Tech University. My research trajectory connects information trust, deep learning evaluation, AI security, and socio-technical systems.

Research Agenda: 2027–2031

Computational Foundations of Trustworthy and Sovereign Machine Intelligence

Guiding question: Can computational trust be formally modelled, learned, calibrated, and integrated into machine learning systems to improve their reliability, robustness, and decision-making under uncertainty?

Pillar 01 · Foundational ML

Trust-Aware Deep Learning

Trust-conditioned representations, uncertainty-aware model adaptation, efficient learning, and robustness under distribution shift.

Research question: When does evidence-aware trust modelling improve on standard confidence and uncertainty baselines?

Pillar 02 · Mathematical Foundations

Computational Trust & Reliability

Probabilistic trust estimation, Bayesian updating, evidence provenance, calibration, and risk-sensitive decision-making.

Research question: How can contextual trust be formally defined and empirically validated?

Pillar 03 · Autonomous AI

Trustworthy Agentic Systems

Reliable tool selection, adaptive delegation, multi-agent collaboration, secure retrieval, and human oversight.

Research question: Can trust-aware policies reduce cascading failures in autonomous workflows?

Pillar 04 · Canadian Deep Tech

Sovereign & Industrial Machine Intelligence

Reproducible evaluation, efficient deployment, data control, and validation in Canadian financial, cybersecurity, and industrial settings.

Research question: How can rigorous ML research strengthen domestic AI capabilities?

This is a developing research agenda. Proposed methods and projects are research objectives, not claims of completed results or secured funding.

Research Projects

Current portfolio work and emerging research directions. Project status and results are documented in the linked repositories.

LLM Evaluation Lab

Reproducible evaluation methods for foundation model performance, reliability, and domain-specific behaviour.

RAG Security

Security evaluation of retrieval-augmented generation, including evidence integrity and adversarial robustness.

Trustworthy AI Evaluation

Evaluation protocols and metrics for trustworthy AI systems.

Information Trust Benchmark

Computational analysis of information credibility and trust-related signals.

Proposed

Trust-Aware Learning Under Distribution Shift

Planned experimental study comparing trust-aware learning against calibration and uncertainty baselines.

Proposed

Trust-Constrained Agentic AI

Planned research into reliable tool selection, evidence assessment, and agent decision-making.

Explore GitHub Research Portfolio ↗

Selected Publications & Scholarly Profiles

My earlier publications establish the theoretical foundations for my continuing work on computational trust and machine intelligence.

Selected works

  • Information trust model (2023)
  • Thinking about Trust: People, Process, and Place (2020)
  • Towards a Computational Model of Information Trust (2018)

See Google Scholar and ORCID for bibliographic records and publication details.

Teaching & Mentorship

My teaching connects cybersecurity, information trust, AI, and information systems with research-led learning and practical evaluation.

Information Trust

Information credibility, computational trust, provenance, and socio-technical systems.

Artificial Intelligence

Machine learning applications, evaluation, and responsible AI development.

Cybersecurity

IT security, biometrics, access control, and security risk.

Contact & Collaborate

I welcome conversations about foundational machine learning, computational trust, trustworthy autonomous AI, graduate research, and Canadian deep-tech collaboration.

Research inquiries

Email: tosatel1@gmail.com

For research collaboration, scholarly correspondence, or speaking inquiries.