INDEX / 01 — 07

Selected work.

Software systems, applied AI, industry data, and experimental research. Each project has its own story and technical context.

THE WORK / 2026
AUTHENTIC PROTOTYPE INTERFACE
ANGE Transport prototype landing interface with navigation, a Rwanda mobility headline and journey-planning actions.
MUHI Mobility — ANGE Transport prototypePrivacy-reviewed application capture · prototype branding
01 / 07
FULL-STACK ENGINEERING · 2026 — PRESENT

MUHI Mobility — Full-Stack Platform

A selective technical case study of a private Python/Flask platform: REST APIs, JavaScript interfaces, SQLAlchemy data modeling and mapping workflows, shown through privacy-reviewed prototype screens.

PythonFlaskJavaScriptREST APIs
EXPLORE PROJECT ↗
FIG. / 02
MOVEMENT AUTHORITY / MESSAGE 3

Illustrative data pipeline · Not a real railway signalling system

CONCEPT / VISUAL STUDY
02 / 07
DATA ENGINEERING · INDUSTRY · JUL — AUG 2026

Structure behind safety.

An industrial internship examining ERTMS/ETCS Message 3, building data dictionaries and prototype validation rules for future V&V work.

PythonPandasOpenPyXLSQL Concepts
EXPLORE PROJECT ↗
FIG. / 03
SMC AI LAB / RESEARCH CONCEPTSYNTHETIC · NO LIVE EXECUTION
ILLUSTRATIVE MARKET CONTEXTSIM / MTF
CURRENT STAGE / 01
01/ 04

Observe.

Synthetic candles and multi-timeframe context are inspected.

ILLUSTRATIVE RECORDResearch observation only

CONCEPTUAL UI · Synthetic candles · No real trading or automatic orders

CONCEPT / VISUAL STUDY
03 / 07
AI RESEARCH · VALIDATION · 2026 — PRESENT

Teaching machines to reason about markets.

An ongoing research initiative exploring how AI-assisted market observations could be grounded in evidence, constrained by deterministic validation and recorded for human review.

PythonStructured DataAI ResearchValidation
EXPLORE PROJECT ↗
LOCAL DEVELOPMENT PROTOTYPE
Corpus Forge local document workspace with upload controls, an empty active corpus and zero documents and indexed words.
Corpus Forge — Local Retrieval StudioActual application capture · local prototype
04 / 07
Python / Document Intelligence / Retrieval-Based Software · 2026

Turning documents into conversations.

A local Flask document workspace for corpus selection, grounded questions, retrieval settings and structured artifacts. A capstone and technical learning foundation that preceded StudyForge AI.

PythonFlaskSQLiteLexical retrieval
EXPLORE PROJECT ↗
AI-RECREATED AWARD ILLUSTRATION
AI-recreated illustration of Trust Layer’s third-place hackathon award, rather than an original diploma scan.
Trust Layer · FINAXYS Build What’s Next 2026Recreated artwork · Original team photograph in the gallery
05 / 07
FINTECH · AI GOVERNANCE · OCT 2026

Trust before action.

An AI governance prototype for payments, KYC/AML and treasury workflows—recognized with third place at FINAXYS Build What’s Next.

PythonLangGraphMulti-agent orchestrationValidation rules
EXPLORE PROJECT ↗
WORKING LOCAL PROTOTYPE
StudyForge student dashboard with course cards, saved materials and an empty practice accuracy indicator.
StudyForge AI — Student Learning WorkspaceActual application capture · local prototype
06 / 07
Applied AI / Education Technology · 2026 — PRESENT

Learning, built around the student.

A student companion connecting course documents, source-backed tutoring and practice, with supporting jobs and document-assistance workflows. An active local prototype, with live AI validation still in progress.

PythonFastAPISQLiteJavaScript
EXPLORE PROJECT ↗
FIG. / 07
CONCEPT / VISUAL STUDY
07 / 07
AI SYSTEMS · SECURITY ARCHITECTURE · 2026 — PRESENT

Private Agentic Enterprise Workspace

Designing a privately deployable, modular AI workspace for authorized tasks, document intelligence and traceable technical workflows. Architecture and feasibility research, with integration and validation still ahead.

Architecture & Feasibility — In Progress

TECHNOLOGIES UNDER EVALUATION
PythonCodexDeepSeek HarnessRAG
EXPLORE PROJECT ↗
THE IMPORTANT DETAIL

Some systems are private, some are prototypes, and some are collaborative. Case studies describe their verified scope without disclosing confidential implementation details.