Greater Melbourne Area
I went from IT support to building production-grade automated intelligence systems — the kind that run at 2am without anyone watching. The path was not traditional. No bootcamp. No team. I taught myself Python, API integrations, concurrent systems design, and core software engineering from scratch — then built something real with it. Something that completes in minutes what a human analyst could not do accurately in an entire lifetime: testing thousands of parameter combinations across 128+ financial instruments, across a decade of historical data, simultaneously. Working independently for 2+ years meant owning every decision: what to build, how to architect it, and how to fix it when it broke at 3am. It built something a job description cannot teach — the discipline to stay consistent without external pressure, and the resilience to push through weeks-long problems with no one to escalate to. The result is Victory101: a production-grade quantitative analytics platform processing 128+ financial instruments across 60+ concurrent processes, with vectorised pipelines delivering a 70%+ reduction in computation time, intelligent DuckDB caching achieving a 99%+ cache hit rate, and five-tier structured logging enabling real-time incident diagnosis. I have since extended the platform into an agentic ecosystem. I integrated LLM-powered pipelines via n8n webhooks to automate live market monitoring — delivering instrument-specific news ingestion directly to Telegram, automated backtest completion notifications, and real-time error alerts with zero manual steps. Throughout development I worked across multiple large language models — GitHub Copilot, Gemini, ChatGPT, and my tool of choice, Claude Code — using each for architecture design, debugging, and iterative code generation, reducing implementation time on complex components by 40%+. The details matter to me. Vectorised operations over loops. Strategic caching over redundant computation. Self-healing retry logic. Fault-tolerant checkpoints. These are not flashy concepts — they are what separate systems that hold under real-world load from systems that do not. I am open to roles across the full spectrum of intelligent systems work: AI and automation engineering is where I want to build, but I am equally at home in data engineering, Python development, systems analysis, and operational or application support where there is room to bring automation thinking into the work. Melbourne-based. Open to remote. Portfolio: github.com/lodson-almeida/python-portfolio
● Machine Learning Lifecycle Implementation: Architected and deployed an end-to-end analytics system encompassing data acquisition, feature engineering, hypothesis validation, and production-grade model evaluation. ● Statistical Validation: Implemented walk-forward analysis to rigorously separate in-sample optimization from out-of-sample validation, reducing overfitting risk and ensuring statistical robustness of model outcomes. ● Data Engineering: Engineered memory‑optimised (float32) data pipelines using Pandas and vectorised NumPy operations to standardise complex multi‑source temporal datasets into deterministic workflows, delivering a 5× increase in analytical throughput. ● Parallelised Experimentation: Established a systematic framework testing thousands of configurations through 60 concurrent processes, utilising checkpoint and recovery systems for fault-tolerant, large-scale validation. ● Production API Deployment: Integrated REST and Streaming APIs with production-grade error handling, retry logic, and five-tier structured logging system (PriceStreamer, Error, Debug, Component-level, Performance logs). ● Data Integrity Framework: Implemented multi-stage data validation and quality assurance checks with automated anomaly detection, ensuring analytical results are derived from verified, high-integrity data sources. ● LLM & Automation Pipeline: Deployed production n8n pipelines integrating LLM-powered news ingestion, webhook triggers, and Telegram notifications, eliminating 100% of manual monitoring steps across live trading operations. Key Achievements: ● 100% workflow Automation: Fully automated the analytical lifecycle, integrating custom Streamlit dashboards and automated Excel engines to deliver KPIs and performance benchmarking. ● Self-directed learner: Demonstrated sustained self-directed learning and technical initiative, translating domain knowledge into production-grade analytical systems with minimal external guidance.
● Cloud Infrastructure: Administered Microsoft 365 and Azure environments, managing user identities, security groups, and access permissions to ensure secure and reliable system connectivity. ● ERP System Support: Provided L1/L2 technical support for Pronto ERP and corporate systems, diagnosing complex software issues and maintaining high-availability for core business applications. ● System Deployment: Orchestrated the provisioning and deployment of corporate hardware and software, ensuring rapid setup and seamless integration for a large-scale user base. Key Achievement: ● Support Workflow Optimization: Redesigned internal SOPs for repetitive technical incidents, reducing first-contact resolution time by approximately 20% through standardized diagnostic procedures.
● Cloud & Identity Administration: Managed user lifecycle and security protocols within Azure, MS Dynamics, and SharePoint; engineered custom SharePoint site templates to streamline digital migration across departments. ● Support Lifecycle Automation: Developed a comprehensive "Bring Your Own Device" (BYOD) connectivity framework and self-help repository, reducing associated support tickets by 80% through user-facing automation. ● Security & Infrastructure Management: Solely administered Barracuda Firewall systems and wireless security portals across international offices (Vanuatu, Solomon Islands), ensuring network integrity and SLA compliance. ● Knowledge Infrastructure: Authored technical SOPs for Intune MDM device enrollment and PowerShell-based automated enrollment scripts, enabling team members to resolve complex mobile management issues. ● Process Optimization: Eliminated recurring manual tasks for "Unit 4 Accelerator" and "Smart Client" software by engineering automated self-service guides, significantly reducing ticket backlogs. ● Leadership & Training: Conducted technical workshops for global stakeholders and C-suite executives on Microsoft Teams and OneDrive, facilitating seamless collaboration for international leadership seminars. Key Achievement: ● Operational Cost Reduction: Reduced mobile hardware repair expenses to negligible levels by identifying display failure patterns, negotiating vendor contracts for protective equipment, and standardizing deployment SOPs.
● User Support: Provided first-line technical support to 500+ users, achieving a 90% first-contact resolution rate through diligent troubleshooting and diagnostic procedure application. Key achievement: ● Knowledge Base Development: Standardized over 20 ‘how-to’ guides and video FAQs, establishing a fundamental self-service process that measurably reduced reliance on manual support.
● Demonstrated strong multitasking skills by shifting smoothly from one task to another. ● Implemented leadership skills through training five new employees over a two years period. ● Acquired customer service skill while serving up to 120+ customers per day. ● Positive can-do attitude with the ability to thrive under pressure, use initiative and be proactive.