AI Solutions Architect
nash@nashtgc.space
Nashiru Muniru, CBAP®, CRISC®
© 2025 NashTGC. All Rights Reserved
Download CV Book a CallIntroduce
Hi, I'm Nashiru Muniru — AI Solutions Architect & AI Systems Engineer
CBAP® · CRISC® · PMI-ACP® · Springer-Published Researcher. I build production AI systems at the intersection of cybersecurity, energy efficiency, and building automation — extending the HVAC-CyberEnergy-CoDesign Framework into cloud-native deployments on AWS, layered with RAG threat co-pilots and agentic incident-response systems.
4
Peer-reviewed
Publications
3
Master's degrees across
CS, Innovation & Engineering
About
I design and ship production AI/ML systems for cyber-physical infrastructure — where security, energy, and operations meet.
My published research (Springer, Cluster Computing, 2026) introduces an AI-based enhanced Twofish encryption scheme using Substitution-and-Shuffle (SnS) and extended cyclic groups, demonstrated on a secure-voting mechanism. I'm now extending the same crypto + AI co-design into a deployable cloud system on AWS for HVAC/building automation — combining a real-time intrusion-detection model trained on RT-IoT2022, a RAG-based threat-explanation co-pilot built over CISA OT/IoT advisories, an agentic incident-response layer using LangGraph, and Streamlit + FastAPI dashboards for operations.
The framing is national-priority: U.S. buildings consume 40% of national energy, HVAC is the largest commercial share, and BAS attacks rose 9× from 2023 to 2024. The work sits squarely inside the National AI Initiative Act (2020) and CISA's OT/IoT cybersecurity priority.
Three Master's degrees (Computer Application Technology · Innovation & Entrepreneurship · Applied Engineering, in progress) and global engagements across North America, Asia, and Africa give me a triangulated view of how AI systems actually need to behave in production.
Featured Research
HVAC-CyberEnergy Co-Design Framework
Co-Optimizing Energy Efficiency and Cybersecurity in AI-Driven HVAC Systems: A Multi-Objective Reinforcement Learning and Lightweight Cryptographic Co-Design Framework for Building Automation.
Theoretical foundation: Constrained Multi-Objective Pareto-Optimal Control for Cyber-Physical Systems — simultaneous optimization of energy cost (J₁) and security cost (J₂).
Springer Publication (Feb 2026)
Muniru & Sun. Cluster Computing. AI-based enhanced Twofish encryption: empowering a secure voting mechanism with SnS and extended cyclic groups. The crypto + AI co-design from this paper is the foundation I'm carrying forward into the HVAC framework. Read on Springer →
TECH 6292 Research Proposal
Full integrative framework (multi-objective RL + crypto co-design). May 2026.
STAT 5784 ML Paper
Machine Learning-Based Intrusion Detection for IoT Networks in Building Automation. RT-IoT2022 dataset (123,117 instances, 83 features), 7 classifiers compared. May 2026.
Cloud-native deployment on AWS
RAG threat co-pilot + agentic incident-response. Public repos: hvac-cyber-id · hvac-threat-copilot · hvac-ir-agent.
Why it matters: U.S. buildings consume 40% of national energy, HVAC is the largest commercial share, and BAS attacks rose 9× from 2023 to 2024. The work sits inside the National AI Initiative Act (2020) and CISA's OT/IoT cybersecurity priority.
Resume
Education, Experience & Publications
Founder & Principal AI Solutions Architect
eLogik.ai · Ontario, Canada — defining solution patterns for RAG, multi-agent orchestration (CrewAI, AutoGen, LangChain), and LLM integration across client AWS and Azure environments. Lifted enterprise AI user satisfaction 30% through systematic defect triage, output validation (precision/recall/F1/ROC-AUC), and CRISC-aligned human-in-the-loop checkpoints aligned to PIPEDA.
Product Manager (AI-Enhanced Platforms)
Appnovation · Toronto, ON — lifted primary conversion 8.5% over six months via structured A/B tests across prompt variants and model choices. Translated stakeholder requirements into AI-ready user stories with explicit model inputs, output expectations, edge cases, and acceptance criteria.
IT / VTMIS Officer
Ghana Maritime Authority — operated and hardened the Vessel Traffic Management Information System (VTMIS), a national-scale OT environment, via access controls, network segmentation, patch cycles, and incident response protecting safety-critical maritime surveillance.
MSc Applied Engineering — Applied Systems Engineering Management
USA — University of Northern Iowa (in progress)
Master of Management in Innovation & Entrepreneurship
Canada — Smith School of Business, Queen's University
MSc Computer Application Technology (thesis-track)
China — Xi'an University of Technology
AI based enhanced Twofish encryption: empowering a secure voting mechanism with SnS and extended cyclic groups
Muniru & Sun (Feb 2026). Cluster Computing, Springer. DOI →
Adaptive Total Variation Image Denoising
IEEE. View →
Region-Fusion Bregman Method for TV Denoising
Springer. View →
Machine Learning for Cybersecurity: PE-File n-grams + Deep Learning for Malware Detection
CBAP® · CRISC® · PMI-ACP® · Lean Six Sigma Green Belt · AWS Cloud Practitioner
Targeting AWS Solutions Architect Associate · July 2026
Advanced Certificate in Applied Data Science with Python
USA — University of Michigan
Advanced Certificate in Computer Security and Systems Management
USA — University of Colorado
Advanced Certificate in Introduction to Cyber Security
USA — New York University
UNI Intellectual Property Award (2025)
University of Northern Iowa
Ghana Innovative Product of the Year (2025)
Awarded to myTapDom.
What I build
Three pillars, one HVAC narrative
Production AI/ML Systems
End-to-end ML pipelines for high-stakes domains — research-backed, containerized, observable. Detection models, classification systems, anomaly engines, with proper evaluation rigor (precision/recall on minority classes, calibration, drift).
Python · PyTorch · scikit-learn · FastAPICyber-Physical & IoT Security
AI for OT environments — building automation, industrial control, IoT fleets. Threat detection, RAG-based threat-explanation, lightweight cryptography for constrained devices (Twofish + Substitution-and-Shuffle networks, per my published work).
RT-IoT2022 · CISA OT advisories · Lightweight cryptoCloud-Native Architecture (AWS)
Reference architectures and production deployments on AWS — SageMaker for serving, Lambda + Step Functions for orchestration, S3 + Athena for telemetry, CloudWatch for ops, KMS for at-rest encryption.
AWS SAA target · July 2026Toolkit
Frequently used tools & frameworks
portfolio
Featured Projects
HVAC Intrusion Detection System — flagship project
Python re-platform of STAT 5784 work. Containerized FastAPI inference on RT-IoT2022 (123,117 instances, 7 classifiers). AWS SageMaker deployment Q3 2026. Repo: hvac-cyber-id.
HVAC Threat Co-pilot
RAG over CISA OT/IoT advisories — explains detected threats in plain language. Repo: hvac-threat-copilot.
HVAC Incident-Response Agent
LangGraph-based agentic loop on top of detection + RAG — drafts and routes IR playbooks with human-in-the-loop approval. Repo: hvac-ir-agent.
contact
Let's build something
I'm currently focused on shipping the production extension of the HVAC-CyberEnergy-CoDesign Framework and pursuing AI Solutions Architect roles. Open to:
- Full-time AI Solutions Architect / AI Systems Engineer roles
- Advisory engagements with cyber-physical / OT-security / building-automation teams
- Speaking & writing on AI for industrial cybersecurity
nash@nashtgc.space
* Marked fields are required to fill.