Loading Samuel Oladejo's portfolio

Samuel Oladejo — Software Engineer, Applied AI

saol@portfolio: ~

$

About me

~/about.md

I'm a software engineer with 4 years of experience building production AI and cloud data-platform systems across FinTech. I run Contour Labs, a startup focused on product development and tech consultancy, taking projects from client scoping to production.

I work full-stack: LLM applications with grounding and verification layers, RAG systems on Google Cloud, ETL pipelines and backend APIs. Before Contour I shipped AI reporting at Xero, and before that built data pipelines at Syft Analytics (acquired by Xero). I'm also finishing an M.Sc. in Computer Science (by Research) at the University of Johannesburg on deep reinforcement learning for portfolio optimisation.

~/skills

Languages

PythonTypeScriptJavaScriptC#SQL

AI / ML

LLM IntegrationGenerative AIPrompt EngineeringLLM EvaluationRAGLangChainVertex AIPyTorchNLPTime-Series Forecasting

Data

ETL PipelinesBatch ProcessingData ModellingPostgreSQLpgvectorSQLitePandasNumPy

Backend / Frontend

FastAPINode.jsREST APIsAsync Job Pipelines.NETReactViteLeaflet

Cloud / Infra

Google Cloud (Cloud Run, Cloud SQL, Vertex AI)AWSDockerCI/CDObservabilityGitUnix/Linux

Consulting

Technical ConsultingStakeholder ManagementClient Delivery

Education

University of Johannesburg

Research: Deep Reinforcement Learning & Financial Mathematics for portfolio optimisation.

Skills: Python, TensorFlow, Financial Maths, Docker, Keras, NumPy, Sklearn, Deep Learning, FIN-RL, Pandas, Matplotlib

University of the Witwatersrand

North-West University

Certifications

CC Certified in Cybersecurity

ISC2 · 2023

credential

Associate Cloud Engineer

Google Cloud · 2023

credential

Solutions Architect Associate

Amazon Web Services · 2022

credential

Cloud Practitioner

Amazon Web Services · 2022

credential

Experience

AI/ML & Software Consulting · est. 2021 · contourlabs.xyz

@@ Client: Confidential (Scientific / Technical Reporting) @@

  • Led delivery from client scoping to production: proposed the RAG architecture and shipped a report-generation app (ingest → outline → drafting → DOCX/PDF export with grounded citations) on FastAPI and React + TypeScript.
  • Designed the retrieval and LLM stack (hybrid TF–IDF + embeddings, domain-aware ranking, streaming with timeout/retry/cancellation) and deployed to GCP (Cloud Run, Cloud SQL + pgvector) with CI/CD and observability.

@@ Client: Confidential (Geospatial / GovTech) @@

  • Architected a production RAG system for natural-language querying over 100K+ geospatial features, combining ChromaDB with a multi-signal retriever (semantic, TF–IDF, fuzzy match, H3 proximity) on FastAPI and GeoPandas.
  • Integrated open-source and cloud LLMs with intent-aware routing and fallback to optimise cost and latency; shipped to Cloud Run with a Leaflet UI rendering GeoJSON.

Projects

~/projects/ai-gis-analyst.py

AI GIS Analyst

Production geospatial RAG.

Led the development of iziMapping, an intelligent geospatial platform for natural-language querying over 100K+ features. Combined ChromaDB with a custom multi-signal retriever (semantic, TF–IDF, fuzzy match, H3 proximity) and intent-aware LLM routing, with a Leaflet UI rendering GeoJSON results.

PythonFastAPIChromaDBGeoPandasShapelyH3LeafletLLMs
~/projects/rag-systems-with-gemma-and-langchain.py

RAG Systems with Gemma & LangChain

Implemented Retrieval-Augmented Generation pipelines.

Implemented embedding, indexing, and retrieval to enhance response accuracy and minimize hallucinations using open-source frameworks.

PythonNumPyLangChainGemmaFastAPIOllamaVertex AIPandas
~/projects/deep-q-networks-and-ppo.py

Deep-Q Networks & PPO

Reinforcement learning for MiniHack environments.

Developed agents using deep-Q networks (3DQN) and Proximal Policy Optimization (PPO) to solve different Mini-Hack Environments with a team of 4.

PythonNumPymatplotlibPyTorchOpenAI gymTensorBoard
~/projects/cnns-for-speech-emotion-detection.py

CNNs for Speech Emotion Detection

Audio analysis for emotion prediction.

Developed a CNN model that converted audio to tensors and predicted human emotions with 84% accuracy.

PythonTensorFlowKerasLibrosamatplotlibsklearnsoundfileSeabornPandas
Code View Project
~/projects/autonomous-surveillance-bot.py

Autonomous Surveillance Bot

AI-driven security robot simulation.

Programmed and deployed a security robot (in simulation) that scans surroundings for suspicious objects and avoids obstacles to reach targets.

PythonRoboticsROSSeabornSLAMMotion Planning

Talks

GDG Devfest Johannesburg 2025 · 01 Nov 2025

Beyond Retrieval: Accelerating AI Development with Google ADK

Shifted the room from RAG to agentic systems, covering multi-agent design, tool use, orchestration, and the ADK build-deploy-evaluate lifecycle.

Impact: 300+ in-person / online

Deep Learning Indaba Kigali 2025 · 22 Aug 2025

Human-AI Interaction in the Global Majority

Co-organized a workshop on oversight patterns for deployed AI systems and building a community repository of real-world human-AI use cases.

Deep Learning IndabaX · 08 Jul 2025

ForecastRL

Presented a compact risk-aware framework for financial trading, combining neural forecasting, reinforcement learning, and portfolio risk modeling.

GDG Devfest Johannesburg 2024 · 09 Nov 2024

Democratizing AI: Building a RAG System with Gemma and LangChain

Covered a practical RAG pipeline and how to deploy it on Cloud Run and Cloud SQL, then extended the session through hands-on GenAI mentoring.

Community

Recent mentorships

Deep Learning Indaba Kigali 2025 · 22 Aug 2025

Human-AI Interaction Workshop

Co-organized a workshop on AI oversight in the Global Majority, documenting real-world use cases and patterns of human decision support.

Co-organizer and mentor · 24 Apr 2025 - 15 May 2025

AI Study Jam

Helped guide a 5-week learning sprint from beginner to intermediate AI/ML practice with hands-on exercises and project support.

Technical mentor · 25 Sep 2025

Google DeepMind AI Hackathon

Mentored teams prototyping AI solutions across HealthTech, EdTech, FinTech, AgTech, and GovTech, with emphasis on architecture and applied Google AI tools.

Hands-on mentor · 09 Nov 2024

Google Cloud Hero labs

Guided participants through Gemini 1.5 Pro labs on Vertex AI, helping them complete challenges and apply GenAI tooling in practice.

Co-led mentor · 27 Jul 2024

Google I/O Extended workshop

Ran a hands-on generative AI lab with Vertex AI and Gemini, while mentoring participants through Google Cloud Skills Boost pathways.

Leadership & Volunteering

Mentor — Google Developer Groups (GDG) Johannesburg

Apr 2024 - Present

Co-Lead (Hackathon / Partnerships) — Deep Learning IndabaX

Feb 2024 - Present

Mentor — Eskom Expo for Young Scientists

Feb 2023 - Feb 2025

Sub-Committee Member (Developer) — Wits Developer Society

Apr 2022 - Dec 2022

Mentor — NEMISA Data Science Hackathon

Feb 2021

Fundraising & RAG Officer — Student Campus Council (SCC), NWU

Aug 2019 - Dec 2020

Awards

Post Graduate Merit Award — University of the Witwatersrand

2021

Golden Key Honour Society Award — North-West University

2018

Affiliations

Black in AIDeep Learning IndabaGoogle Developer GroupIEEEIITPSA

Get in touch

~/contact/mail.shlive

Send an Email

Drop me a message directly to my inbox. I usually respond within 24-48 hours.

Email
~/contact/calendar.shlive

Schedule a Meeting

Book a 30-minute slot directly on my calendar.

> connection established · press Ctrl/⌘ K to jump anywhere