Based in Lagos, Nigeria

Moshood
Alimi

AI Application Engineer

I build production AI systems that solve real problems.

I work at the intersection of software engineering and AI — building RAG systems, agents, and LLM-powered products that ship.

About

I'm a software engineer with a full-stack JavaScript background who moved into AI engineering by building — RAG pipelines, vector databases, agents, fine-tuned models, and data pipelines, all shipped to production.

I contract with Blue AI, an EdTech platform using GPT-4o and OCR to automate student script marking. I've also independently shipped two products: a personal finance analyzer for Nigerian bank statements with a custom fine-tuned categorization model, and a football scouting platform with a config-driven player rating engine calibrated on 10,000+ match samples.

I'm looking for remote AI engineering roles where I can keep building things that work.

RAGLLMsGPT-4oTypeScriptPythonQdrantLangChainMCPNode.jsNext.jsEmbeddingsFunction Calling

Projects

Blue AI

Production App

Production OCR + GPT-4o pipeline that scans handwritten student scripts and marks them against teacher-uploaded marking schemes. Human-in-the-loop review before AI processing.

GPT-4oOCRPythonReactRAG

Finance Analyzer

Production App

Full-stack personal finance platform for Nigerian bank statements. Upload statements from OPay, Kuda, UBA, Zenith, or GTBank — the system auto-detects the bank, extracts every transaction, and classifies it using a hybrid pipeline: a custom fine-tuned HuggingFace model handles common transactions, a Groq LLM covers ambiguous ones, and a second verification pass catches misclassifications. Dashboard shows spending breakdown, monthly cash flow, top recipients, income sources, and AI-generated monthly insights.

PythonFastAPIReactTypeScriptHuggingFaceGroqSupabaseRecharts

KnowBall

Football scouting platform that rates players from raw match data across five position groups — strikers, wingers, CAMs, midfielders, and defenders. Each position has a config-driven rating engine scoring 7 dimensions, calibrated against 10,000+ match samples. Players are ranked against positional peers with league and cross-league comparisons. Includes role archetype classification, style profiling, and auto-generated scout report cards. Daily data pipeline pulls from Sofascore and Understat via GitHub Actions.

PythonReact 19TypeScriptTanStackSupabasePostgresGitHub Actions

LangChain Football Tactical Agent

Multi-step agent that scrapes Sofascore with Playwright, combines match stats with web search, and generates tactical analysis for Premier League fixtures.

LangChainPlaywrightTypeScriptAgents

Qdrant Knowledge Base

Persistent semantic search system supporting markdown, PDFs, and URLs. Metadata filtering, dynamic vector dimensions, deployed with live URL.

QdrantTypeScriptEmbeddingsVector DB

Writing

My First AI Engineering Mock InterviewLinkedIn

A reflection on preparing for AI engineering roles and what the process revealed about the field.

I Learned Transformers by Arguing With an AI. Here's the Transcript.Substack

This is a conversation between Claude and me where we learn about transformers and how attention mechanisms work, step by step.

Evaluation Metrics for LLMs and RAG SystemsLinkedIn

Why Evaluation is Hard?

Most 'AI Engineers' Can't Answer These Basic QuestionsSubstack

This is a Q&A between Claude and me to gauge my understanding of this book I read titled AI Engineering by Chip Huyen.

Read more on Substack →

Let's work together

I'm open to remote AI engineering roles and freelance projects. Best way to reach me is email.

moshood988@gmail.com