AI Demo and Portfolio

AI Demos and Portfolio

Examples of my AI-assisted development work, AI-enabled solution designs and working applications can be viewed by clicking on the VIEW DEMO links below.

Bringing enterprise governance to AI-assisted software development

Screenshot below, I have built a web-based platform around Amazon Web Services' emerging AI-Driven Development Lifecycle (AI-DLC) methodology that transforms AI-assisted software development into a structured, governed process suitable for enterprise environments. The platform walks through every step of the AI SDLC and enables business analysts, project managers and business stakeholders to collaborate alongside AI by providing: Guided requirements discovery User story management Human approval gates Live project progress AI governance and traceability Automated Word documentation Visibility into AI-generated design decisions Rather than replacing software engineering processes, the platform demonstrates how AI can strengthen governance, improve collaboration and accelerate delivery while keeping humans in control.

Independent Well Integrity Management System (WIMS)

Independently designed and developed a working Well Integrity Management System (WIMS) using Claude Code, translating detailed engineering, operational and regulatory business requirements into a functioning application. The project demonstrates the ability to move rapidly from business requirements and solution design through AI-assisted development to working software, applying the same enterprise thinking used within complex Oil & Gas environments. The solution demonstrates key Well Integrity capabilities including risk classification, integrity assessments, failure and action management, regulatory workflows, auditability and operational decision support, while exploring the application of AI agents, intelligent workflow orchestration, anomaly detection, predictive analytics and LLM-assisted decision support within a safety-critical engineering context.
View WIMS Demo

Lumen — AI Business Intelligence Platform

Designed, built and deployed Lumen, an AI-enabled business intelligence application demonstrating how generative AI can provide business leaders with a more intuitive interface to operational data. The application combines structured business data with an LLM-powered conversational layer, enabling users to interrogate information in natural language and receive contextualised insights rather than relying solely on conventional dashboards and reports. Built using React, Vite, Tailwind CSS and the Anthropic Claude API, Lumen includes an AI-generated executive brief and contextual AI workspaces across Marketing, Operations, Customer and Executive functions. The application was taken from concept to deployed solution in approximately five days using Claude Code, demonstrating rapid AI-assisted prototyping and delivery of business-facing applications.
View BI Platform Demo

Mortgage Freedom Tracker — Commercial Mobile Application

Independently designed, developed and launched Mortgage Freedom Tracker, a commercial consumer mobile application released through the Google Play Store. Using Claude Code as an AI-assisted development partner, I took the product from concept through architecture, development, testing and production release in approximately two weeks. The application incorporates a sophisticated mortgage forecasting and repayment calculation engine, interactive dashboards, repayment and journey tracking, notifications, onboarding, gamification mechanics, milestone celebrations and commercial monetisation through in-app purchasing. The project was intentionally undertaken to demonstrate the productivity gains achievable through modern AI-assisted software engineering. By comparison, while leading technology at Lite n’ Easy, I engaged an external software development agency to deliver a significantly less feature-rich mobile application that required approximately six months and more than A$120,000 to develop. Mortgage Freedom Tracker demonstrates how experienced technology leaders can combine domain expertise, architecture judgement and AI-assisted development tools to dramatically compress delivery timelines and reduce the cost of taking software from concept to production. The working application can be downloaded by searching on the Play Store for Mortgage Freedom Tracker or clicking the link below
Mortgage Freedom Tracker Mobile App

Senior Technology & Data Transformation AI Focused Consultant / Analyst

Contract Engagement

Technology leader with a strong foundation in data integration and operational systems within complex Oil & Gas environments, focused on driving AI-enabled transformation initiatives that improve efficiency, data quality, and risk management. Experienced in bridging business and technology to design and implement AI-enabled and agent-driven workflows, aligning emerging capabilities with enterprise priorities, governance requirements, and measurable business outcomes.

This portfolio demonstrates practical applications of AI in real-world operational contexts that I designed, with emphasis on solution design, workflow integration, and business value delivery, leveraging modern AI platforms and cloud ecosystems (e.g. Amazon Bedrock on Web Services, Microsoft Azure AI, Google Vertex AI).

  • Led end-to-end current-state assessment of Well Integrity Management System (WIMS) across the full well lifecycle, identifying systemic gaps in data governance, workflow design, and regulatory assurance within a safety-critical environment.
  • Engaged cross-functional stakeholders (Production Engineering, Asset Services, Technical Authority, Field Operations) to map and validate integrity workflows, uncovering heavy reliance on manual processes, fragmented data sources, and low system trust.
  • Identified critical limitations in existing WIMS implementation, including inability to support automated risk classification (traffic light model), lack of auditability, and absence of real-time integrity monitoring across thousands of wells.
  • Designed a target-state AI-enabled Well Integrity Monitoring & Assessment solution, leveraging machine learning–based anomaly detection and time-series analysis to identify early indicators of well integrity risk.
  • Defined architecture incorporating AI agents, LLM-assisted workflows, and retrieval-augmented data access (RAG) to dynamically generate context-aware assessments, automate data ingestion and validation, and orchestrate workflows across engineering teams.
  • Introduced concept of agent-based workflow orchestration, enabling automated task routing, natural language explanations of anomalies, and intelligent prepopulation of integrity assessments—reducing manual effort and improving decision consistency.
  • Highlighted integration strategy across enterprise systems (SAP, MDM, Nexus, OCIS) using API-led architecture to enable unified data access and support AI-driven analytics and regulatory workflows.
  • Reframed WIMS program from a system upgrade to an enterprise-wide integrity management transformation, positioning AI as a key enabler for proactive risk management, predictive analytics, and scalable compliance.
  • Developed comprehensive business and regulatory requirements covering workflow automation, integrity assessment, failure management, and advanced analytics, forming the basis for future-state platform evaluation and solution design.
  • Delivered actionable insights to senior stakeholders, influencing strategic direction toward AI-enabled decision support, improved auditability, and data-driven integrity assurance.
AI-ENABLED WELL INTEGRITY RISK MONITORING - DREW Salem

USE CASE 1: AI-ENABLED WELL INTEGRITY RISK MONITORING

Problem

The Well Integrity Management System relied on fragmented and inconsistent data sources, resulting in:

 MASP and pressure data distributed across multiple systems

  • Data inconsistencies, latency, and misalignment
  • Heavy reliance on manual interpretation by engineers
  • Missed early warning signals and leading indicators
  • Disjointed, manual workflows across stakeholders

 Additionally:

 Well Integrity Assessments were conducted using static, spreadsheet-based forms, often including irrelevant fields

  • Assessment results required manual analysis, increasing time, effort, and risk of human error
  • Risk calculations (current and future) were manually derived
  • Follow-on workflows (regulatory, remediation, close-out) were manually triggered and managed

 The AI-Enabled Solution

Designed an AI-assisted and agent-driven Well Integrity workflow that enhanced pre-assessment, assessment, and post-assessment processes through automation, intelligent analysis, and dynamic workflow orchestration.

 The solution integrated:

  • AI-driven data conditioning and anomaly detection
  • Agent-based workflow orchestration
  • Dynamic digital assessment forms
  • AI-assisted risk modelling and decision support

  

Solution Architecture Overview

The solution was structured across two core phases:

Phase 1: AI-Driven Pre-Assessment & Early Risk Detection

 Workflow Design

  • Ingest pressure and well data from multiple logs, databases, and operational systems via APIs
  • AI layer performs:
    • Data segregation to isolate relevant datasets (e.g. filtering non-relevant scouting log data)
    • Data quality validation and cleansing based on defined rules and patterns
  • Apply AI/ML models and LLM-assisted logic to:
    • Detect early anomalies and leading indicators
    • Identify deviations from expected operational patterns
  • AI generates:
    • Natural language explanations of anomalies
    • Contextual insights using historical data (RAG pattern)

  

Agent-Based Orchestration

AI agents coordinate:

  • Routing anomalies to the appropriate well / reservoir / production engineer based on issue type
  • Triggering alerts and initiating workflows
  • Assigning tasks to operators and engineers for additional data collection
  • Tracking actions and maintaining audit logs

Dynamic Assessment Preparation

  • Generate a context-aware digital Well Integrity Assessment form:
    • dynamically tailored to the identified anomaly
    • irrelevant fields and sections removed
  • Prepopulate form fields with:
    • cleansed data
    • AI-derived insights
    • contextual historical information

 Business Outcome (Phase 1)

  • Early detection of integrity risks
  • Reduced manual data preparation
  • Improved workflow coordination across stakeholders
  • Faster initiation of assessment processes

 

Phase 2: AI-Assisted Core Well Integrity Assessment

 Workflow Design

  • Ingest results from pre-assessment workflows (field inputs, engineer feedback, operational data)
  • AI layer performs:
    • Natural language analysis of submitted data and observations
    • Anomaly detection across combined datasets
    • Pattern recognition across historical and current data

 AI - Driven Risk Assessment

  • Generate:
    • Risk scoring and classification / updating traffic light system
    • Time-series based failure predictions (e.g. estimated time to integrity failure)
    • Identification of contributing factors and risk drivers

Agent-Enabled Decision Support & Workflow Execution


AI agents:

  • Populate and refine the digital Well Integrity Assessment
  • Provide AI-generated recommendations and insights 
  • Trigger:
    • remediation workflows
    • regulatory compliance processes
    • reporting and audit requirements
  • Enable human-in-the-loop validation by engineers prior to finalisation

 

Business Outcome (Phase 2)

  • Reduced assessment time and effort
  • Improved accuracy and consistency of risk evaluation
  • Proactive identification of future integrity risks
  • Streamlined regulatory and remediation workflows

Concepts, capabilities and technologies demonstrated

 Indicative Technologies / Concepts

  • LLM platforms
  • AI agents for workflow orchestra
  • AI agents for workflow orchestration and task management
  • Retrieval-Augmented Generation (RAG) for contextual insights
  • Time-series anomaly detection models
  • API-based integration with engineering systems (e.g. WellView Well Management)
  • Vector databases for semantic search and historical analysis
  • Model Context Protocol (MCP) concepts for controlled data and tool access

 

Key Capabilities Demonstrated

  • End-to-end AI-enabled workflow design (pre-assessment → assessment → remediation)

    • Agent-based orchestration of complex operational processes

  • Integration of AI into existing engineering and data systems
  • Dynamic, context-aware user interface design (adaptive assessment forms)
  • Application of AI to risk modelling and decision support in high-consequence environments

 

Workflow Design

  • Aggregate pressure and well data from multiple systems
  • AI/agent layer:
    • detects anomalies in time-series data
    • retrieves historical context via APIs
    • generates natural language explanations using LLMs
  • Agent orchestrates:
    • alert generation
    • routing to engineers
    • follow-up actions
  • Human-in-the-loop validation and decision logging

 

Indicative Technologies / Concepts Used

  • LLM platforms
  • AI agents for workflow orchestration and alerting
  • Time-series anomaly detection
  • API integration with engineering systems
  • Model Context Protocol (MCP) concepts for controlled data/tool access

 

Business Value Achieved

  • Earlier identification of integrity risks
  • Reduced manual analysis effort
  • Improved decision-making speed and consistency

 

Key Considerations Demonstrated

  • Data quality and consistency
  • False positives and alert fatigue
  • Requirement for human validation