Selected essays and technical reflections.
Reflecting on the challenges, lessons, and personal growth of the past years, and embracing new opportunities, fatherhood, and uncharted horizons in 2025 and beyond.
One confirmed feature currently in editorial review, plus the ideas I'm shaping for future pieces — grid planning, systems strategy, and the human side of engineering.
Tracing the evolution of a sample size from an unjustified default of 500 to a full-census evaluation of 16,676 events.
State-aware screening doesn't just get steadily better; its value is conditional on how far conditions sit past the local violation onset.
How pointed methodology questions from a co-author who never touched the code led to a complete overhaul of the validation pipeline.
A year of real German TSO redispatch instructions reveals a sharp, real geographic structure that perfectly mirrors the Energiewende.
Building systems is never just technical. It’s about how layers connect. A reflection on careers, system design, and building resilience at the interfaces of grids and code.
Seventh entry in this series, and the start of a new thread: does the same audit discipline that found operating state dominates over event magnitude in German redispatch data hold up somewhere structurally different — Australian rooftop-solar-driven overvoltage instead of Ger...
Eighth entry in this series. The previous post ruled out single-customer and network-topology explanations for why rooftop-solar overvoltage wasn't showing up. This post is where the right network was found — and where the same "real data produces zero violations" surprise fro...
Ninth entry in this series. The previous post traced zero violations at every customer count back to the real explanation: 2012–2013 capacity data is too small relative to today's typical installs, not too few customers. This post is where capacity, not count, actually got swe...
Tenth entry in this series, and the one the whole Australia thread was actually building toward. The previous post found perfect recall and collapsing precision past onset — the opposite failure mode from Germany. This post explains why, and it's not a subtle difference. It's ...
A suggested shortcut combined correlation coefficients into a decision threshold. Rejecting it and computing a real finite-difference sensitivity vector doubled screening precision from 1.26% to 2.51% while preserving 100% recall.
AEMO's 67 GW connection queue tests whether event magnitude or operating state governs transmission voltage. Across 960 AC power-flow solves, magnitude dominates (r=0.51 vs r=0.09) while seasonal demand swings leave violation rates completely flat.
Transmission hosting capacity is not a monolith. On the 330/500 kV EHV backbone, data centres cause 0.0% violations up to 2 GW. On 132 kV sub-transmission, 66.4% violate, with 55.3% suffering physical nose-curve voltage collapse.
Because transmission lines are inductive, dV/dQ dominates dV/dP. Operating data centres with leading power factor or adding a 300 MVAr STATCOM expands connection headroom by 2.4x to 3.5x, while uncoordinated clustering in Sydney and Melbourne triggers corridor collapse.
Subspace Identification (SSID) and Kalman filtering deliver 58.5–92.5 min Warning Lead Time on LV feeders and 249–267 min on transmission data-centre ramps. An EarlyWarningTransformer sequence model drops false alarms by 30 percentage points to 15.0% with 100% recall.
Operating state dominates at 20 kV MV, magnitude dominates at 415 V LV, and magnitude dominates at 132–500 kV transmission while impedance dictates feasibility. Synthesizing three real studies into a unified 5-factor decision framework for power systems engineers.
In the Pipeline — Ideas I'm Shaping
How redispatch reality reshaped my thinking about product prioritization.
What parenting can teach engineers about patience, resilience, and better systems.
How lessons from SaaS pipelines are reshaping how I think about 11kV networks and back again.
How smart automation habits from SaaS work help me design cleaner, safer 11kV systems.
Reflecting on the challenges, lessons, and personal growth of the past years, and embracing new opportunities, fatherhood, and uncharted horizons in 2025 and beyond.
A deep dive into the engineering, the dead-ends, and the architecture behind processing grid streams in real-time. A 16-part technical journey spanning German medium-voltage networks and Australian rooftop solar.
A streaming voltage-screening pipeline finds hundreds of violations on its default operating-state table, and exactly zero on real load data. Here is why.
A state-aware filtering rule caught 35 out of 35 known voltage violations. On closer inspection, it turned out to be circularly constructed.
Germany's DSOs publish updated grid expansion plans on October 31, 2026. Here's what §14d EnWG actually requires, why 800+ fragmented DSOs make connections take up to 18 months, and why this window closes in 2028.
Inside my CWIEME Berlin 2026 Future Engineers Programme talk: grid bottlenecks, NSGA-II charging-infrastructure optimization, and why closing the industry's talent gap starts with engineers who bridge hardware, software, and policy.
How we scaled client-side numerical solvers to accelerate VDE-AR-N 4110/4120 grid connection audits. Awarded 2nd Place at CWIEME Berlin 2026.
Smart Grid strength and short-circuit power are the silent constraints shaping HVDC integration.
Thermal bottlenecks, §14a compliance, and PowerFactory simulations reshape equitable EV rollout decisions.
A Python-driven digital twin for Power Grid N-1 analysis and scalable scenario planning.
Why voltage stability is the next big constraint for high-renewable grids, and what it means for planners.
How system design choices can reinforce inequity — and how engineers can counteract it.
Charging equity is not just fair — it is operationally superior.
Geospatial optimization reframed how I think about fairness in EV infrastructure.
From a single question to an IEEE Best Paper — what the community taught me about impact.
A short interview on grid reliability, digital twins, and why engineering must bridge physics and software.
Reflecting on the 'Making More Health' course by Ashoka and integrating systems thinking into health tech product operations.
Examining systemic gender inequality in healthcare access and the foundational steps needed for inclusive systems.
Reflections on transitioning from viewing leadership as authority to understanding it as empowerment and change management.
An examination of how energy, land, labor, and capital impact income redistribution and Kenya's Big Four Agenda.
Examining the impact of World Bank and IMF structural adjustment programmes on Kenya's energy sector and economy.
I also share shorter reflections on LinkedIn, including in-progress ideas on engineering, systems thinking, and technical leadership.
Read on LinkedInOccasional emails when I publish new essays on grid planning, systems strategy, and applied intelligence.
No spam. Unsubscribe anytime. Unsubscribe
I work across power systems, software, and strategy to build solutions that are technically strong and operationally useful.
Start a conversationSelected essays and technical reflections.
Reflecting on the challenges, lessons, and personal growth of the past years, and embracing new opportunities, fatherhood, and uncharted horizons in 2025 and beyond.
One confirmed feature currently in editorial review, plus the ideas I'm shaping for future pieces — grid planning, systems strategy, and the human side of engineering.
Tracing the evolution of a sample size from an unjustified default of 500 to a full-census evaluation of 16,676 events.
State-aware screening doesn't just get steadily better; its value is conditional on how far conditions sit past the local violation onset.
How pointed methodology questions from a co-author who never touched the code led to a complete overhaul of the validation pipeline.
A year of real German TSO redispatch instructions reveals a sharp, real geographic structure that perfectly mirrors the Energiewende.
Building systems is never just technical. It’s about how layers connect. A reflection on careers, system design, and building resilience at the interfaces of grids and code.
Seventh entry in this series, and the start of a new thread: does the same audit discipline that found operating state dominates over event magnitude in German redispatch data hold up somewhere structurally different — Australian rooftop-solar-driven overvoltage instead of Ger...
Eighth entry in this series. The previous post ruled out single-customer and network-topology explanations for why rooftop-solar overvoltage wasn't showing up. This post is where the right network was found — and where the same "real data produces zero violations" surprise fro...
Ninth entry in this series. The previous post traced zero violations at every customer count back to the real explanation: 2012–2013 capacity data is too small relative to today's typical installs, not too few customers. This post is where capacity, not count, actually got swe...
Tenth entry in this series, and the one the whole Australia thread was actually building toward. The previous post found perfect recall and collapsing precision past onset — the opposite failure mode from Germany. This post explains why, and it's not a subtle difference. It's ...
A suggested shortcut combined correlation coefficients into a decision threshold. Rejecting it and computing a real finite-difference sensitivity vector doubled screening precision from 1.26% to 2.51% while preserving 100% recall.
AEMO's 67 GW connection queue tests whether event magnitude or operating state governs transmission voltage. Across 960 AC power-flow solves, magnitude dominates (r=0.51 vs r=0.09) while seasonal demand swings leave violation rates completely flat.
Transmission hosting capacity is not a monolith. On the 330/500 kV EHV backbone, data centres cause 0.0% violations up to 2 GW. On 132 kV sub-transmission, 66.4% violate, with 55.3% suffering physical nose-curve voltage collapse.
Because transmission lines are inductive, dV/dQ dominates dV/dP. Operating data centres with leading power factor or adding a 300 MVAr STATCOM expands connection headroom by 2.4x to 3.5x, while uncoordinated clustering in Sydney and Melbourne triggers corridor collapse.
Subspace Identification (SSID) and Kalman filtering deliver 58.5–92.5 min Warning Lead Time on LV feeders and 249–267 min on transmission data-centre ramps. An EarlyWarningTransformer sequence model drops false alarms by 30 percentage points to 15.0% with 100% recall.
Operating state dominates at 20 kV MV, magnitude dominates at 415 V LV, and magnitude dominates at 132–500 kV transmission while impedance dictates feasibility. Synthesizing three real studies into a unified 5-factor decision framework for power systems engineers.
In the Pipeline — Ideas I'm Shaping
How redispatch reality reshaped my thinking about product prioritization.
What parenting can teach engineers about patience, resilience, and better systems.
How lessons from SaaS pipelines are reshaping how I think about 11kV networks and back again.
How smart automation habits from SaaS work help me design cleaner, safer 11kV systems.
Reflecting on the challenges, lessons, and personal growth of the past years, and embracing new opportunities, fatherhood, and uncharted horizons in 2025 and beyond.
A deep dive into the engineering, the dead-ends, and the architecture behind processing grid streams in real-time. A 16-part technical journey spanning German medium-voltage networks and Australian rooftop solar.
A streaming voltage-screening pipeline finds hundreds of violations on its default operating-state table, and exactly zero on real load data. Here is why.
A state-aware filtering rule caught 35 out of 35 known voltage violations. On closer inspection, it turned out to be circularly constructed.
Germany's DSOs publish updated grid expansion plans on October 31, 2026. Here's what §14d EnWG actually requires, why 800+ fragmented DSOs make connections take up to 18 months, and why this window closes in 2028.
Inside my CWIEME Berlin 2026 Future Engineers Programme talk: grid bottlenecks, NSGA-II charging-infrastructure optimization, and why closing the industry's talent gap starts with engineers who bridge hardware, software, and policy.
How we scaled client-side numerical solvers to accelerate VDE-AR-N 4110/4120 grid connection audits. Awarded 2nd Place at CWIEME Berlin 2026.
Smart Grid strength and short-circuit power are the silent constraints shaping HVDC integration.
Thermal bottlenecks, §14a compliance, and PowerFactory simulations reshape equitable EV rollout decisions.
A Python-driven digital twin for Power Grid N-1 analysis and scalable scenario planning.
Why voltage stability is the next big constraint for high-renewable grids, and what it means for planners.
How system design choices can reinforce inequity — and how engineers can counteract it.
Charging equity is not just fair — it is operationally superior.
Geospatial optimization reframed how I think about fairness in EV infrastructure.
From a single question to an IEEE Best Paper — what the community taught me about impact.
A short interview on grid reliability, digital twins, and why engineering must bridge physics and software.
Reflecting on the 'Making More Health' course by Ashoka and integrating systems thinking into health tech product operations.
Examining systemic gender inequality in healthcare access and the foundational steps needed for inclusive systems.
Reflections on transitioning from viewing leadership as authority to understanding it as empowerment and change management.
An examination of how energy, land, labor, and capital impact income redistribution and Kenya's Big Four Agenda.
Examining the impact of World Bank and IMF structural adjustment programmes on Kenya's energy sector and economy.
I also share shorter reflections on LinkedIn, including in-progress ideas on engineering, systems thinking, and technical leadership.
Read on LinkedInOccasional emails when I publish new essays on grid planning, systems strategy, and applied intelligence.
No spam. Unsubscribe anytime. Unsubscribe
I work across power systems, software, and strategy to build solutions that are technically strong and operationally useful.
Start a conversation
