KZDI AI MCP Pipeline
Summary
Designed and deployed a 7-stage AI safety architecture utilizing Cloudflare Workers for production deployment.
About
Highly innovative AI Systems Builder and Backend Engineer with expertise in NLP pipelines, bilingual AI infrastructure, and cybersecurity-hardened systems for underserved communities. As Founder of KZDI Technologies Ltd, I excel at deploying impactful solutions that transform technology consumers into value creators, driving strategic execution and leveraging deep tech for social and economic advancement. My focus is on creating secure, scalable, and intelligent systems that bridge technological gaps and foster community empowerment.
Work
KZDI Technologies Ltd
|Founder & AI Systems Architect
Northern Nigeria, Nigeria
→
Summary
Led KZDI Technologies Ltd as Founder & AI Systems Architect, driving innovation in NLP pipelines, bilingual AI infrastructure, and cybersecurity solutions for underserved communities in Northern Nigeria.
Highlights
Architected and deployed a 7-stage AI safety MCP pipeline using Cloudflare Workers, implementing robust rate limiting, CORS enforcement, input sanitization, and error handling to enhance system security and reliability.
Developed a bilingual WhatsApp AI Quiz Bot (Hausa/English) leveraging Express.js, Twilio, Railway, and Anthropic API, enabling a 30-minute AI Mentor session mode for enhanced user engagement.
Established a Hausa NLP Pipeline, overseeing the annotation of a 50-sample dataset in Label Studio v1.0, and resolving P1 errors during the pre-training stage to ensure data quality and model performance.
Conducted an ISO 27001 production security audit across 14 domains, completing 49% of a 108-step checklist, specifically applying cybersecurity best practices to enhance system resilience.
Pioneered the PICTDA BPO Pilot, deploying Label Studio live on HuggingFace Spaces and developing an NDPR-compliant government submission pipeline, streamlining data annotation processes.
Launched TabibAI, a doctor-booking platform featuring Islamic UX and Hausa AI triage, implementing a multi-stream revenue model encompassing subscriptions, commissions, and AI upsells.
Engineered a content automation pipeline integrating LunarCrush, Claude API, Notion, Gmail, and Twilio SMS, significantly reducing manual content creation effort.
Education
3MTT Nigeria Cohort 4
Nigeria
Professional Program
Software Engineering, Cybersecurity, Natural Language Processing
Grade: Assessments: 10/10
Courses
Software Engineering Track
Cybersecurity NextGen
DeepTech_Ready: NLP
NECO
Nigeria
Secondary Education
Science & Technical subjects
Grade: 8 Credits
Languages
Hausa
English
Certificates
Microsoft / Digital Skills Nigeria Certifications
Issued By
Microsoft / Digital Skills Nigeria
ISO 27001 Security Audit Practice (Applied)
Issued By
KZDI Technologies Ltd
Cloudflare Workers Deployment & Hardening
Issued By
Cloudflare
Label Studio Annotation Infrastructure (Applied)
Issued By
Label Studio
Skills
Programming Languages & Backend Development
Python, JavaScript, TypeScript, Node.js, Express.js, REST APIs, Backend Development.
Cloud & Infrastructure
Vercel, Cloudflare Workers, Railway, Supabase, PostgreSQL, Git/GitHub.
AI & Natural Language Processing
Prompt Engineering, Intent Classification, Dataset Annotation, Hugging Face, Label Studio, MCP Architecture, NLP Pipeline Design, Bilingual AI Products.
Cybersecurity & Compliance
ISO 27001, Threat Modeling, Rate Limiting, CORS Hardening, Input Validation, Auth Systems, API Security Hardening.
Business Operations & Strategy
Technical Documentation, Business Analysis, Systems Thinking, Digital Strategy, AI Workflow Architecture, Community Digital Education, Digital Transformation.
Projects
Hausa NLP Pipeline
Summary
Developed an intent classification system for Hausa NLP, utilizing HuggingFace and Label Studio with a 50-sample annotated dataset.
WhatsApp AI Quiz Bot
Summary
Developed a bilingual quiz engine with Node.js/Railway and AI Mentor mode, with Supabase integration planned.
PICTDA BPO Pilot
Summary
Implemented an annotation infrastructure using HuggingFace Spaces, compliant with NDPR, for a government pilot project.