5+ years in applied AI/ML
Olatoye Olaide Gabriel
AI/ML Engineer
I train, evaluate, and deploy machine learning systems: LLM assistants with RAG and tool calling, computer vision, and time-series models. My AI tutors run in production learning platforms with hundreds of active learners.
- Python
- PyTorch
- TensorFlow
- scikit-learn
- XGBoost
- LangChain
- OpenAI API
- FastAPI
Selected work
Project index
Sigmant
Sigmant is an AI-powered algorithmic and quantitative trading platform with 17 skills and 123 agent tools. I built it with Next.js and TypeScript for the dashboard and API layer, Python FastAPI for the skill services, XGBoost, LightGBM, and PyTorch for the models, MongoDB, Redis, and Parquet for data, and Socket.IO for live updates. Orders pass risk checks and trader confirmation before reaching one of 7 brokers.

CorralGPT (NUC Course Parser)
CorralGPT is an AI course-writing workspace that turns Nigerian NUC curriculum PDFs into editable course records. I built it with FastAPI (62 endpoints) and SQLAlchemy (23 models) for the backend, PyMuPDF for rule-based parsing, five LLM provider adapters streaming drafts over SSE, and React with Tiptap for an editor that exports to Word.

KBM LMS Platform
KBM LMS is a Moodle 4.5 learning platform with 32 custom plugins serving 200+ active learners. I built it with PHP plugins adding 93 database tables, an AI tutor that calls OpenAI, Claude, or Gemini through a tool-calling loop, Paystack, Flutterwave, Stripe, and HitchPay checkout, a Word-to-course importer, and Redis, Nginx, and Bash for operations.

MLGuard
MLGuard is a network intrusion detection prototype that classifies live traffic. I built it with Scapy to capture packets into bidirectional flows with 68 features, scikit-learn scaling and 50-component PCA, an XGBoost classifier for attack, normal, and scanning traffic, and Flask with Socket.IO streaming results to a React dashboard with CSV export.
Skills & tools
Languages
Python for ML pipelines and FastAPI services. TypeScript and JavaScript for React, Next.js, and Node.js applications. PHP for Moodle plugins and Laravel. Rust for the InkWell desktop app.
See the workAI & machine learning
Trained and evaluated XGBoost, CNN, and PPO models. Built RAG and tool-calling assistants on OpenAI, Gemini, and Ollama, and served live model inference for intrusion detection.
See the workBackend & data
Designed REST and WebSocket APIs, database-per-tenant isolation in Laravel and MySQL, atomic order claims in MongoDB, and payment settlement that verifies before it writes.
See the workPlatform & delivery
Run production Moodle platforms on Linux: release pipelines, Redis caching, CDN video, SSL, health checks, and daily backups. Integrated Paystack, Flutterwave, and SeerBit payments.
See the workFrontend & mobile
Built dashboards, student portals, marketing sites with a CMS page builder, and an offline-first React Native attendance app.
See the workQuant & research
Built a Gymnasium market simulator, trained a PPO policy, and evaluated signals on time-ordered splits with fees, slippage, and matched random baselines.
See the work
Project index
Case studies, prototypes, and research.
Visit Sigmant (opens in a new tab)Flagship project · Algorithmic trading platform
Sigmant
Sigmant is an AI-powered algorithmic and quantitative trading platform with 17 skills and 123 agent tools. I built it with Next.js and TypeScript for the dashboard and API layer, Python FastAPI for the skill services, XGBoost, LightGBM, and PyTorch for the models, MongoDB, Redis, and Parquet for data, and Socket.IO for live updates. Orders pass risk checks and trader confirmation before reaching one of 7 brokers.
LLM tool calling · Purged walk-forward validation · Risk and execution controls · Microservices
View CorralGPT screenshotsAI / ML
CorralGPT (NUC Course Parser)
CorralGPT is an AI course-writing workspace that turns Nigerian NUC curriculum PDFs into editable course records. I built it with FastAPI (62 endpoints) and SQLAlchemy (23 models) for the backend, PyMuPDF for rule-based parsing, five LLM provider adapters streaming drafts over SSE, and React with Tiptap for an editor that exports to Word.
Rule-based PDF parsing · Provider adapters · Streaming over SSE
Visit KBM Learning (opens in a new tab)LMS Platform
KBM LMS Platform
KBM LMS is a Moodle 4.5 learning platform with 32 custom plugins serving 200+ active learners. I built it with PHP plugins adding 93 database tables, an AI tutor that calls OpenAI, Claude, or Gemini through a tool-calling loop, Paystack, Flutterwave, Stripe, and HitchPay checkout, a Word-to-course importer, and Redis, Nginx, and Bash for operations.
Moodle plugins · Webhook verification · Permission-aware tool calls
Explore MLGuard on GitHub (opens in a new tab)AI / ML
MLGuard
MLGuard is a network intrusion detection prototype that classifies live traffic. I built it with Scapy to capture packets into bidirectional flows with 68 features, scikit-learn scaling and 50-component PCA, an XGBoost classifier for attack, normal, and scanning traffic, and Flask with Socket.IO streaming results to a React dashboard with CSV export.
Bidirectional flow aggregation · PCA · Live model inference
More in this field
Tayari Tutor Matching
Tayari Tutor Matching recommends tutors to language learners on the Tayari marketplace. I first trained a Keras scoring model on 300 synthetic student-tutor pairs, then rebuilt matching as semantic search: tutor profiles are embedded locally with Ollama nomic-embed-text in batches of 16 and stored in LanceDB, and learner requests are searched with SQL-style filters for budget, subjects and time-slot overlap. I then ported it into the team marketplace as an Express tutor-matching service.
Cassava Disease Classifier
I built a cassava leaf disease classifier that uses a frozen ImageNet ResNet50 as a feature extractor and a 100-tree Random Forest on the pooled features, trained on 7,508 leaf images across 5 classes. The hybrid reached 0.89 test accuracy against 0.59 for a Random Forest on raw pixels, and I serve it through a Streamlit app with upload, camera and sample-image tabs.
BrainWave
BrainWave is an online learning platform where students enrol in tutor-led courses, join live classes and use an AI assistant. I built the TypeScript Express backend with 88 routes across 16 route modules and 15 Mongoose models, Stripe and wallet payments, Jitsi JaaS tokens for live sessions and Socket.IO messaging. Its AI service routes chat to Gemini, OpenAI or Anthropic from a MongoDB setting and adds context from 6 skill files. I also built the student app and a separate admin console.
Tayari AI Course Media
Tayari AI Course Media generates curriculum, narration and video for tutors building language courses on the Tayari marketplace. Inside the team's Express backend I wrote provider classes for ElevenLabs and Google text-to-speech and for Veo 2 video. The Veo provider renders up to 8 clips from distinct, culturally specific scene prompts, joins them with FFmpeg and merges the narration track. Seven instructor-only routes expose outline, lesson, exercise, audio, video, metadata and lecture generation.
AFLON Smart Campus Concierge
The Smart Campus Concierge lets staff upload school handbooks and policies and ask questions by text or voice. I built it as a client-side React app in about 1,100 lines: PDF.js extracts text, 4,000-character chunks with 1,000-character overlap are stored in IndexedDB, keyword scoring picks up to 30 passages plus neighbours, and Gemini 2.5 Pro streams the answer. Voice uses the Gemini Live native-audio model.
Hospital AI Translation
I built a translation prototype for doctor-patient conversations: a Flask API serving three fine-tuned MarianMT models that I published on Hugging Face (Yoruba to English, Nupe to English, Nupe to Yoruba), and a Flutter client that posts text to the API. The repo includes a 144-sentence trilingual clinical phrase dataset.
Academic research
I supported postgraduate research in intrusion detection, health informatics, and decision intelligence by designing and implementing machine learning models.
Real-Time Intrusion Detection System Using PCA and Ensemble Machine Learning
Impact by organization
Nigeria
Know the Blocks Maven (KBM)
8 production web properties · 200+ active learners
Led delivery across Moodle, websites and enrolment funnels. Built the course tutor and Paystack enrolment flow.
Wamido
2 checkout currencies
Led a separate Moodle deployment and added Flutterwave payment verification in NGN and USD before enrolment.
ACAIR
500+ active users on CNS Courseware
Led development of the Moodle LMS and built its release pipeline. Shipped separate releases for the ACAIR student portal, administration app, and API.
LAUTECH
2020–2025
Developed ML models for postgraduate research and co-authored a paper on intrusion detection using PCA and ensemble models.
Global Tech Computer Institute
100+ learners taught
Taught HTML, CSS, JavaScript and PHP through projects and mentored learners in Lagos.
Kenya
Tayari.Live
200 tutor profiles indexed
Built semantic tutor matching with Ollama embeddings and LanceDB, and AI course media with ElevenLabs, Google TTS, and Veo video joined by FFmpeg.
Remote
TeraAI
2024–2025
Built trading dashboards and market-data APIs and developed a PPO research simulator with position accounting and slippage.

About
I build machine learning into products people use.
AI/ML engineer with 5+ years of experience training, evaluating, and deploying machine learning systems across NLP, computer vision, reinforcement learning, and retrieval-augmented generation (RAG). I take models from research to production and have the full stack and DevOps experience to ship them myself. In 2025 and 2026 I put AI tutors into Moodle learning platforms that serve 200+ active learners at KBM and 500+ active users at CNS Courseware.
- Location
- Abuja, Nigeria · open to remote work
- Experience
- 5+ years in AI/ML · 7 years in software
- Current roles
- AI/ML Engineer & Platform Lead, KBM · Manager of AI/ML, ACAIR
- Education
- B.Sc. Computer Science, First Class Honours · LAUTECH
- Research
- 5 co-authored papers in ML, cybersecurity, and blockchain
Experience
Know the Blocks Maven (KBM)
AI/ML Engineer & Platform Lead
2026–Present · Remote
Training organisation delivering AI, blockchain, and Web3 programmes across Nigeria.
I lead engineering at KBM across product, code, infrastructure, and the developers on the team. I designed and shipped 8 production web properties, centred on the KBM learning platform: Moodle 4.5 extended with 32 custom plugins (about 132,000 lines of PHP, 93 database tables, 19 scheduled tasks) serving 200+ active learners. It includes an AI tutor on OpenAI, Claude, and Gemini, four payment gateways, gamification with verifiable certificates, parent accounts, sprint cohorts, learner analytics, and Nigeria Data Protection Act compliance. I also operate it in production and replicated it as a second, isolated tenant for Wamido.
- Built a multi-provider AI tutor (OpenAI, Claude, Gemini) as a bounded tool-calling agent with four tools: browse courses, recommend courses, remember, and read learner progress. Course cards always come from tool results, and it serves 200+ active learners across KBM and Wamido.
- Gave the tutor per-learner memory that learners can view and erase, and locked its lesson access to what the learner can open after an audit found course IDs could be walked to paraphrase paid content.
- Built Doc-to-Course, a deterministic OOXML parser that maps Word headings to Moodle sections, labels, and pages, keeps tables, lists, and images, and re-imports idempotently.
- Built predictive learner analytics with a students-at-risk view, and an AI-assisted lead-generation and re-engagement system for a community of 500+ members.
- Mentor the developers on the team, set up code review and CI/CD, and wrote and taught AI and blockchain course curricula.
- Delivered 8 production web properties end to end: two Moodle learning platforms, two corporate websites, a partner site, a founder portfolio, and two enrolment funnels.
- Extended Moodle 4.5 with 32 custom plugins adding 93 database tables and 19 scheduled tasks: catalogue and checkout, a branded theme with 38 templates, workbooks, forum, XP, streaks, badges and verifiable certificates, parent accounts, sprint cohorts, analytics, and NDPA compliance.
- Built four payment gateways: Paystack, Flutterwave, Stripe, and HitchPay. Paystack settlement locks the race between browser return and signed webhook, re-verifies the stored reference and amount, supports saved-card recurring charges, and refuses test keys on the live host.
- Moodle 4.5
- PHP 8.3
- JavaScript
- Mustache
- SCSS
- OpenAI API
- Anthropic
- Gemini
- Paystack
- Flutterwave
- Stripe
- MariaDB
- Redis
- Nginx
- Bash
- Next.js
- Sanity
- MongoDB
- Resend
Sigmant
Independent AI/ML & Full Stack Engineering
Independent project
My own AI trading platform for researching markets, replaying strategies, and managing trades within risk limits.
I designed and built Sigmant end to end: an AI-powered algorithmic and quantitative trading platform with 17 Python FastAPI skill services, a Next.js dashboard, and a conversational trading agent with 123 typed tools. It covers market data, a 76-feature model pipeline, purged walk-forward validation, risk controls, and execution across 7 brokers, with 4,500+ automated tests.
- Built a conversational trading agent with 123 typed tools across 17 FastAPI skill services. Read tools run at once; tools that place or change orders only return a proposal the trader must confirm.
- Built time-ordered model evaluation that purges overlapping labels between chronological splits, and benchmarked signals against matched random entries on 27,622 usable rows under the same fees and slippage.
- Built the Next.js dashboard and 160 API route files that trace a decision from signal to order status, with 4,500+ automated test cases across pytest and Jest.
- Built a market replay that accounts for overlapping positions, fees, and slippage, and used XGBoost move estimates to screen trades against round-trip costs.
- Added a conditional MongoDB update so only one request can claim an approved proposal, and checked account risk limits before each broker dispatch.
- Python
- FastAPI
- XGBoost
- LightGBM
- PyTorch
- Next.js
- TypeScript
- MongoDB
- Redis
- Socket.IO
- Docker
ACAIR (Aflon Centre for AI & Robotics)
Manager of AI/ML & Lead Senior Software Engineer
2025–Present · Abuja, Nigeria
EdTech and AI research organisation building digital learning and smart-campus platforms.
I lead AI/ML and software at ACAIR, an applied AI and robotics institute, and mentor its junior engineers. I architected and shipped its learning and campus platforms: the cns.com.ng Moodle courseware platform with 500+ active users, the ACAIR student portal, admin app, and payments API, the CorralGPT curriculum parser and AI writing workspace, and the aflon.edu.ng smart-campus assistant.
- Built the NUC Course Parser behind CorralGPT. It extracts Nigerian university curriculum PDFs into editable course records with PyMuPDF, then streams model drafts from the records an educator selects.
- Architected the aflon.edu.ng smart-campus AI concierge with Next.js, Gemini APIs, RAG, and a vector database. The platform has 415 students enrolled.
- Built a facial-recognition attendance system with a CNN, TensorFlow, and OpenCV to verify student identity in real time.
- Integrated OpenAI and Ollama learning assistants into production Moodle platforms, including cns.com.ng.
- Led and mentored junior engineers across model training, evaluation, deployment, and code review.
- Moodle 4.5
- PHP
- Next.js
- React
- React Native
- Node.js
- Express
- MongoDB
- Python
- FastAPI
- TensorFlow
- OpenCV
- Gemini API
- Bash
TeraAI
Full Stack & AI/ML Engineer
2024–2025 · Remote
Fintech startup building automated trading infrastructure.
I built AI and full stack features for TeraAI’s automated trading products: reinforcement learning research on a custom Gymnasium market simulator, signal models, and the Node.js APIs and React dashboards traders used.
- Built a Gymnasium market simulator with position accounting and slippage, then trained and backtested a Stable-Baselines3 PPO trading policy on historical data.
- Developed signal models that combine reinforcement learning, multi-modal LSTM, and fusion attention for financial signal generation.
- Designed Node.js, Express, and MongoDB APIs for market predictions, analytics dashboards, wallet-to-wallet transactions, and automated trade execution.
- Built the React, Next.js, and Tailwind CSS dashboards traders used to follow model output in real time.
- Python
- Gymnasium
- Stable-Baselines3
- PPO
- Node.js
- Express
- MongoDB
- React
- Next.js
- Tailwind CSS
Tayari.Live
AI/ML & Backend Engineer
2024–2025 · Remote · Kenya
EdTech platform connecting learners with tutors across East Africa.
I built the AI services for Tayari’s African-language tutor marketplace: semantic tutor matching on local embeddings and an AI course-media pipeline that generates curriculum, narration, and video.
- Built tutor matching for an African-language tutor marketplace: first a Keras ranking model, then semantic search that embeds 200 tutor profiles with Ollama nomic-embed-text in LanceDB and filters by budget, subjects, and time-slot overlap.
- Built a course-content generation service on LLM APIs.
- Developed Node.js and TypeScript APIs that expose the recommendation and content services to the platform frontend.
- TypeScript
- Node.js
- LangChain
- Ollama
- Vector database
- RAG
LAUTECH
Academic Researcher
2020–2025 · Ogbomoso, Nigeria
Ladoke Akintola University of Technology, Cybersecurity and AI research.
- Designed and implemented machine learning models for intrusion detection, health informatics, and decision intelligence across Master’s and PhD research projects.
- Co-authored five papers on machine learning, cybersecurity, and blockchain, including a study of real-time intrusion detection with PCA, Random Forest, and XGBoost.
- Python
- scikit-learn
- XGBoost
- PCA
- SMOTE
- Solidity
- IPFS
Global Tech Computer Institute
Web Development Instructor
2019–2020 · Lagos, Nigeria
Computer training institute.
- Taught HTML, CSS, JavaScript, and PHP to 100+ learners through a project-based curriculum, and mentored them from beginner to advanced level.
- HTML
- CSS
- JavaScript
- PHP
Education
B.Sc. Computer Science
Ladoke Akintola University of Technology (LAUTECH)
2020–2025 · Ogbomoso, Nigeria
First Class Honours
Coursework and research in machine learning, algorithms, and software engineering.
Research and publications
Certificates

Applied AI/ML for Professionals
Aflon Digital Institute · ACAIR
20 April 2026

Applied AI and Advanced Robotics for Professionals
Aflon Digital Institute · ACAIR
20 April 2026

Certificate of Publication · Real-Time Intrusion Detection
Advance Scholars Publication
16 April 2026

NDG Linux Unhatched
Cisco Networking Academy
7 October 2020
