BEST MAZHINDU
ACADEMIC CURRICULUM VITAE
Prospective Doctoral Applicant — AI Governance, Evaluation & Safety
Zimbabwe │ Remote GMT+2 │ Open to international opportunities
best@bestmazhindu.com · +263 77 552 6709 · linkedin.com/in/best-mazhindu · github.com/Best-Mazhindu · bestmazhindu.com
RESEARCH PROFILE
Prospective doctoral researcher in AI Governance, Evaluation & Safety, bringing 8+ years of enterprise and cloud architecture experience across security, governance, technology risk and AI-enabled transformation. I am currently Tech Lead at Analytics Hive, where I lead enterprise cloud and security architecture and advise enterprise and public-sector clients, including on secure AI adoption. I hold an MBA in AI & Business Strategy and a BSc in Information Technology.
My emerging research interests concern how organisations govern, evaluate, secure and responsibly deploy AI systems across their lifecycle, with attention to accountability, transparency, human oversight and risk management, and particular interest in regulated and resource-constrained environments in developing-country contexts. Years of governing complex technology environments, securing systems and advising on technology decisions give me a practitioner, systems-level perspective on these problems, which I now wish to investigate through formal doctoral research.
This CV reflects applied industry experience and independent, self-directed projects. It does not claim peer-reviewed publications, a formal research appointment or prior academic research output.
RESEARCH INTERESTS
Primary research area: AI Governance, Evaluation & Safety
•AI Governance — organisational structures, policies, accountability mechanisms, risk management and lifecycle governance for AI systems.
•AI Evaluation & Auditing — methods for evaluating AI systems for safety, reliability, robustness, transparency, fairness and governance readiness.
•AI Safety & Responsible Deployment — practical mechanisms for identifying, mitigating and monitoring AI risks before and after deployment.
•Trustworthy & Responsible AI — operationalising principles such as accountability, transparency, human oversight, privacy, security, fairness and technical robustness.
•Enterprise & Public-Sector AI Governance — governance approaches for deploying AI in regulated, high-risk and resource-constrained organisational environments, with particular interest in developing-country contexts.
PROSPECTIVE DOCTORAL STUDY — APPLICATION PHASE
Intended field: Artificial Intelligence
Proposed research area: AI Governance, Evaluation & Safety
Not currently enrolled. This represents an intended doctoral research trajectory and is subject to admission, funding and supervisory arrangements. Target institution and programme to be confirmed.
EDUCATION
2022MBA, Artificial Intelligence & Business Strategy
Zimbabwe Open University
•Studied machine learning strategy alongside enterprise transformation, responsible AI governance and executive digital leadership.
2015BSc, Information Technology
Midlands State University
•Foundation in algorithms, data structures, software engineering and information systems (see Awards).
PROFESSIONAL EXPERIENCE
Industry roles are presented as the practical foundation for doctoral inquiry into AI governance. They are professional work, not academic research.
Jan 2026 – PresentTech Lead | Enterprise & Cloud Architect
Analytics Hive — international data, analytics and IT consultancy · South Africa, UK, Zimbabwe
•Lead enterprise cloud and security architecture for enterprise and public-sector clients, embedding IAM least privilege, network security and security-by-design into solution designs aligned to the AWS Well-Architected Framework.
•Provide technical direction to a cross-functional team of 7 across architecture, cloud engineering and data/AI; run architecture reviews for security, scalability and cost.
•Advise Heads of IT, Risk and Operations on cloud, security and AI transformation, and shape architecture governance standards and delivery methodology.
•Structure cloud, security, data and secure AI/ML capabilities into enterprise offerings. This work on secure AI adoption informs my emerging interest in how organisations govern and evaluate AI systems.
2021 – 2025AWS Cloud Architect
Poshi Creative — cloud consultancy · SMEs, schools and microfinance · Southern Africa
•Architected and operated a multi-tenant AWS environment supporting 50+ live production applications across 15+ client accounts, giving sustained exposure to technology governance, risk, security and accountability in complex production systems.
•Standardised IAM governance across client accounts (role hierarchies, enforced MFA, quarterly audits), reducing over-privileged access by over 60%.
•Designed layered VPC architectures with subnet segmentation and security-group policy to limit cross-environment exposure.
•Established monitoring and audit logging (CloudWatch with SNS, CloudTrail) across 15+ environments, and built reusable Terraform/CloudFormation modules.
•Advised 10+ clients on migration strategy and redesigned cost architecture for three major clients, achieving 20–25% reductions without degrading SLAs.
Oct 2016 – Dec 2020Software & Systems Engineer
IS Systems Solutions — financial and commercial clients
•Delivered four multi-system enterprise integration projects, building grounding in enterprise integration and financial-sector systems.
APPLIED RESEARCH & AI GOVERNANCE-RELEVANT ACTIVITIES
Applied, practitioner-led and self-directed activities. None is presented as peer-reviewed research.
•AI & Emerging Digital Academy (AEDA), Founder — established a training and AI-literacy initiative; partnered with Leaderman to design AI capability-building programmes and to examine human–AI collaboration in Zimbabwean organisations. Applied and practitioner-led; not peer-reviewed.
•“AI at Work” Workshop, Kumathom Media Solutions (26 Nov 2025) — designed and delivered a full-day workshop for managers and leaders from banking, retail, healthcare, NGO and public-sector organisations, covering applied AI concepts and prompting. Relevant to AI literacy and organisational AI adoption; a dissemination activity, not research.
•Applied Research Project — AI Verification & Decentralised Trust (MVPV), Lead — exploratory applied research into AI verification layers integrated with blockchain architecture. Exploratory and non-peer-reviewed; a secondary technical interest rather than a research direction.
•Intelli-Sys, Founder — AI solutions practice delivering AI chatbot integration, cloud deployment and automation services for client web platforms. Professional practice.
SELECTED PROJECTS
Independent / self-directed portfolio projects unless otherwise noted. No experimental results are claimed beyond those stated.
Secure AI & Data Architecture (featured)
•Secure AI & Data Architecture — portfolio/reference architecture, not a production deployment. Designs governance and security requirements into an AI platform using Amazon Bedrock and SageMaker on S3 with data lifecycle policies, IAM least privilege, KMS encryption, Bedrock Guardrails, and Responsible AI and privacy requirements. It illustrates how governance requirements can be translated into concrete architectural controls, and is the project closest to my proposed research direction.
AI Deployment & Governance
•Enterprise RAG Chatbot — LangChain and GPT-4 retrieval-augmented generation pipeline over SharePoint content, with semantic chunking and vector retrieval.
•SageMaker & REST API Pipeline — trained classification and NLP models with SageMaker and served inference through FastAPI endpoints in Docker.
AI Systems & Evaluation
•NLP Fraud Detection Engine — fine-tuned BERT and Scikit-learn ensemble models, containerised to AWS Lambda for real-time inference.
•Customer Churn Predictor — XGBoost classification model with SHAP-based explainability.
RESEARCH & TECHNICAL CAPABILITIES
•AI Evaluation & Governance: AI system evaluation · AI governance · Responsible AI · risk assessment · AI lifecycle governance · security-by-design · auditability (audit logging, access reviews).
•AI/ML: NLP · Generative AI · LLM/RAG systems · model deployment · Python experimentation · explainability (SHAP).
•Systems: AWS (IAM, KMS, VPC, CloudTrail, CloudWatch, Bedrock, SageMaker) · cloud architecture · Docker · Terraform · Kubernetes · REST APIs.
CERTIFICATIONS & PROFESSIONAL DEVELOPMENT
Presented as technical preparation for research in AI governance, secure AI systems and enterprise technology governance, not as academic credentials.
•Certified: AWS Solutions Architect – Professional (SAP-C02) · AWS Solutions Architect – Associate (SAA-C03) · AWS Security – Specialty.
•In progress: CISSP · TOGAF (Enterprise Architecture) · DPO / Data Protection.
•Further development: NIST AI RMF and ISO/NIST governance frameworks (self-study) · Responsible AI and ITIL 4 badges.
AWARDS & DISTINCTIONS
•16 distinctions in Information Systems coursework — Midlands State University, 2015.
•Top Student, Muzinda Hub Developer Community — Zimbabwean developer-training programme; mentored peers in Python, machine learning and cloud development.
PROPOSED RESEARCH TRAJECTORY
Emerging research direction. The questions below are potential areas of inquiry, not confirmed doctoral research questions, and all plans are subject to admission, funding and supervision.
Overarching question: How can organisations design effective governance and evaluation mechanisms that enable the safe, accountable and trustworthy deployment of AI systems throughout their lifecycle?
Potential research questions
•How can AI governance principles be operationalised into organisational processes?
•How can AI systems be evaluated for safety, reliability and governance readiness, and how can evaluation and auditing mechanisms support accountability?
•How should human oversight be incorporated into high-impact AI systems?
•How can governance frameworks adapt to rapidly changing generative and agentic AI systems?
•How can technical AI evaluation be integrated with risk, security, compliance and governance processes?
•How can these approaches be adapted to developing-country and African organisational contexts?
Research Proposal Sketch
Proposed pilot (not yet conducted; no results claimed): Evaluating the effectiveness of guardrail controls, such as Amazon Bedrock Guardrails, against prompt injection in enterprise retrieval-augmented generation (RAG) systems deployed in regulated African financial-services settings, using the NIST AI Risk Management Framework as the evaluation structure. The study would (1) build a reference RAG system and a documented set of prompt-injection test cases; (2) measure guardrail performance, including detection, false-positive and bypass rates; (3) map findings to the NIST AI RMF functions (Govern, Map, Measure, Manage) to assess governance readiness; and (4) examine, through practitioner interviews with risk and IT leaders, how such evidence could support accountability, human oversight and audit within regional regulatory expectations, and where governance gaps persist in practice. The intended contribution is a practical, reproducible evaluation method that resource-constrained organisations could adapt to their own AI deployments. Scope, methods and ethics approval would be refined with a supervisor.
Near term: apply for doctoral study in Artificial Intelligence with a focus on AI governance, evaluation and safety, bringing 8+ years of industry experience to formal academic inquiry.
Longer term: develop original research and publish in peer-reviewed venues; contribute to a future African AI research community.
LEADERSHIP & ADVISORY EXPERIENCE
Architecture governance · technical leadership · stakeholder and client advisory · architecture decision-making. Applied technical and leadership experience across 30+ organisations, including:
•National IT Director, TJG — national-level technology leadership, IT strategy, governance and infrastructure direction.
•Cloud Consultant, Greenmasters / Ridgeview Investments — cloud architecture, infrastructure and technical advisory.
•AI Automation Engineer, MorpheLabs — applied AI automation solutions.
LANGUAGES
English — fluent · Shona — fluent
REFERENCES
Academic referee: Prof. [Name, title, institution, email] — MBA supervisor
Professional referee: [Name, title, email] — Analytics Hive