Electrical Engineer to AI Engineer: A Real Story and Practical Transition Roadmap

An Electrical Engineer's Career Crossroads

On Reddit, an electrical engineer from Brazil shared his thoughts on a potential career transition — and it resonated with many. His story is remarkably relatable: 33 years old, six years of experience in energy distribution, a 2019 electrical engineering degree, currently working as an engineer at a major power transmission company, earning around BRL 15k/month pre-tax, with meal allowances, profit-sharing (PLR), and solid health insurance.

About PLR (Profit Sharing): PLR (Participação nos Lucros e Resultados) is an employee incentive mechanism under Brazil's labor law framework, regulated by Law No. 10,101 of 1994. Companies are required to distribute a portion of annual profits to employees, and this income enjoys preferential treatment at the individual income tax level. For state-owned or semi-public enterprises in energy and telecom, PLR can be quite substantial — sometimes equivalent to one to three months of base salary. It's a key component of Brazil's formal employment benefits system and a hidden benefit this engineer must factor in when weighing the cost of leaving.

From a traditional engineering standpoint, this is an enviable position. The power sector offers stable pay, low layoff risk, and a clear path to retirement. Yet he found himself conflicted: should he give up this stability to become an AI engineer?

His core desire wasn't actually about salary — it was about freedom in how he works. The energy sector pays well, but the work culture remains traditional. He currently works from home only two days a week, while what he truly wants is a fully remote lifestyle.

Remote Work and the Structural Shift in Engineering Careers: After 2020, remote work evolved from a perk to a standard feature of many tech roles — but this trend has played out very differently across industries. AI and software development positions offer significantly higher remote rates than traditional infrastructure fields. According to Levels.fyi and LinkedIn data, roughly 40–55% of AI engineering roles in the US and Europe offer fully remote or hybrid options, compared to less than 15% for traditional energy engineers. For engineers in South America, going fully remote opens up another possibility: accessing a global labor market priced in USD or EUR while living on local currency costs — a meaningful geo-arbitrage opportunity. In 2024, the Brazilian real held at roughly 5–6 BRL per USD for extended periods, meaning a Brazilian engineer with a dollar-denominated remote role could enjoy purchasing power four to five times greater than a local equivalent salary. This economic logic is profoundly reshaping how Latin American tech talent thinks about career mobility, and it's a key driver fueling demand for cross-border remote hiring platforms like Toptal and Remote.com.

From Distribution Analyst to Self-Taught Data Science

This engineer didn't start from scratch with programming. His technical journey follows a clear trajectory:

At ENEL SP (São Paulo): Joined the "losses" department as a junior analyst, developing predictive algorithms to identify potential electricity theft — and falling in love with data science in the process.
Self-study phase: Rather than pursuing a graduate degree, he taught himself Python and machine learning through books, YouTube, and ChatGPT.
At Equatorial (Goiás): Worked on one of the most critical processes in the distribution sector — indicator accounting — applying his ML knowledge to develop algorithms that delivered measurable, quantifiable results across the entire franchise area covering seven of Brazil's distribution companies.

Brazil's Energy Sector Context: Brazil is Latin America's largest electricity market, with a grid dominated by hydropower (roughly 60% of installed capacity) and regulated by the national energy regulator ANEEL. ENEL SP and Equatorial are among Brazil's major electricity distributors, serving São Paulo state and Goiás state respectively. The issue of "losses" (perdas) — both technical and non-technical (i.e., electricity theft) — has long plagued Brazilian distributors, costing billions of reais annually. This is precisely why predictive algorithms and ML have direct commercial value in the sector.

The Natural Advantage of Electrical Engineers Transitioning to AI: The overlap between electrical engineering and machine learning is far deeper than most people realize. Signal processing, control systems, and time series analysis are core EE curriculum topics — and they also happen to be foundational algorithmic domains in industrial AI. More importantly, electrical engineers are trained to translate physical-world constraints (such as voltage limits and power balance) into mathematical models. This ability to "formalize engineering problems" is exactly what many ML engineers from pure CS backgrounds lack. Power systems data also exhibits classic multivariate time-series characteristics with strong physical constraints, meaning engineers who've worked with such data tend to adapt quickly to other industrial AI contexts — at far lower transition cost than non-STEM career changers.

The Strategic Value of a Portfolio for AI Transitions: In the AI engineering job market, portfolios now systematically outweigh academic credentials as a signal of capability — a trend that has accelerated sharply since 2022. The reason: AI ability is hard to verify through traditional degree signals, while open-source projects on GitHub, Kaggle competition rankings, model releases on Hugging Face, or a well-written technical blog post that clearly explains project context and quantified outcomes are far more persuasive than a diploma. For engineers with real business experience, the most competitive portfolio content often comes from "sanitizing and retelling" actual past projects — preserving the methodology and quantitative results while stripping out commercially sensitive details. The indicator accounting algorithms this engineer built at Equatorial, properly framed, would make a highly compelling portfolio piece. The phrase "delivered measurable gains across seven distribution companies" is itself a strong recruitment signal.

Reddit original post screenshot

This experience reveals an important truth: he was already doing work closely resembling AI/data science within a traditional engineering role. This "transition" isn't a leap across a chasm from zero — it's more about getting proper credit for skills he already has.

Weighing Stability Against Freedom

He openly acknowledges that staying in the energy sector means guaranteed income and a comfortable path to retirement in a sector that's largely insulated from layoffs. But he raised a deeper question: is it worth it in the long run — especially from a mental health perspective?

This touches on an anxiety shared by many in tech today.

The Hidden Cost of Stability

Stability in traditional industries often comes bundled with rigidity: mandatory on-site presence, strict hierarchies, and slow-moving change. For someone who already has modern technical capabilities and craves a flexible lifestyle, this environment can be a slow drain. He even said he'd be willing to accept a pay cut in exchange for a fully remote role.

Mental Health and the Systemic Problem of Burnout: Engineer burnout has received growing academic and industry attention in recent years. According to a 2023 Blind survey of tech workers, engineers in traditional infrastructure sectors (energy, telecom, manufacturing) reported significantly higher rates of "career stagnation" than those in internet and AI companies. The core triggers weren't salary-related — they were "limited growth opportunities" and "lack of autonomy in how work gets done." Psychology researchers define this state as Chronic Mild Burnout — distinct from acute emotional exhaustion, it's characterized by persistently low motivation, low engagement, and a creeping sense of meaninglessness in one's work. It often appears invisible from the outside while causing ongoing psychological damage to the individual. This hidden cost is frequently overlooked in career decision models, yet for driven technical professionals, its long-term impact can far outweigh short-term salary considerations.

The Practical Realities of Transitioning

He posed several key questions to the community:

Is the AI engineering field highly competitive right now?
What does a realistic transition path actually look like?
For someone with some programming background, what should a transition roadmap look like?

Understanding the AI Engineer Career Path: AI Engineer is an emerging role that has differentiated itself from data scientist and machine learning engineer positions in recent years. The core responsibility is integrating large language models (LLMs) or machine learning models into real products and systems. Unlike research-oriented ML scientists, AI engineers emphasize engineering delivery — including API integration, fine-tuning, RAG (Retrieval-Augmented Generation) architecture design, vector database management, and production deployment. This role's rise is directly tied to the generative AI explosion post-2022, as enterprises urgently need engineers who can "plug AI capabilities into business workflows" — and genuinely well-rounded talent combining domain knowledge with engineering ability remains scarce.

RAG Architecture and Enterprise AI Deployment: RAG (Retrieval-Augmented Generation) is one of the core architectural paradigms for enterprise AI implementation today. Proposed by Meta AI in 2020 and rapidly adopted as standard engineering practice after ChatGPT went mainstream, its basic principle is: before querying a large language model, first retrieve relevant document chunks from an external knowledge base (typically stored as a vector database), inject them as context into the prompt, then have the LLM generate a response. This architecture effectively mitigates LLM "hallucination" and knowledge cutoff limitations, allowing companies to give models access to proprietary domain knowledge without expensive fine-tuning. For power sector AI applications, RAG can be used to build intelligent equipment fault diagnosis assistants, regulatory compliance Q&A systems, or grid operations knowledge bases — exactly the kinds of scenarios where engineers with domain expertise can quickly create value.

AI Transition Tech Stack for Engineers: For engineers with Python foundations and ML experience, the skill-completion path to becoming an AI engineer is relatively clear. The first layer is MLOps and engineering capability — MLOps (Machine Learning Operations) is the key engineering discipline bridging model development and production deployment, borrowing DevOps philosophy to bring CI/CD culture into ML lifecycle management. This includes Docker containerization, FastAPI service deployment, model monitoring (MLflow, Weights & Biases), and cloud platforms (AWS SageMaker, GCP Vertex AI). For industrial AI scenarios like power forecasting, MLOps is especially critical because grid load patterns shift continuously with seasons, economic cycles, and consumption structure — without solid monitoring and retraining pipelines, model accuracy degrades significantly within months. The second layer is generative AI engineering — frameworks like LangChain or LlamaIndex, vector databases (Pinecone, Weaviate), prompt engineering, and RAG system design. The third layer is vertical domain integration — applying these capabilities to specific scenarios like power forecasting, equipment fault diagnosis, and intelligent scheduling. For engineers who already have scikit-learn experience and time-series data processing skills, the first layer typically requires just three to six months of focused study to reach interview-ready level, while the third layer's competitive advantage requires almost no additional learning to establish.

Practical Advice for Engineers with Similar Backgrounds

Based on this engineer's actual profile, the following observations may be useful for others in similar situations.

Underestimated Core Advantages

Many people assume transitioning to AI means starting from scratch — but this engineer already has: real-world ML deployment experience in actual business contexts, Python programming ability, and project outcomes demonstrating "using algorithms to solve practical problems and generate quantifiable results." The combination of domain knowledge and ML capability is precisely what engineers with purely AI backgrounds lack. Demand for AI applications in energy, power, and industrial sectors is growing rapidly. Industry experience isn't a liability — it's a differentiating competitive advantage.

Transition Doesn't Have to Be Radical: Three Stable Paths

For a 33-year-old with family considerations and a need for stability, a full quit-and-leap may not be the optimal move:

Seek out more AI/data-related projects within the current role, building a portfolio of work that can be shown externally;
Target energy tech (Energy Tech) or industrial AI companies — these firms need AI capability but also value industry background, and they tend to be more open to remote work;
Make "fully remote" a hard filter when evaluating target roles, rather than broadly applying to all AI positions.

An Honest Assessment of Competition

Entry-level AI engineering roles are genuinely competitive. But what's truly scarce has never been people who can call APIs — it's always been the well-rounded AI talent that "understands the business, can ship, and has real engineering chops."

Competitive Landscape in Vertical AI: Unlike the intense competition for generalist AI roles, AI applications in energy, power, and industrial verticals are in a phase of rapid penetration with relatively scarce talent. The IEA's 2023 report noted that AI applications in grid optimization, demand forecasting, predictive equipment maintenance, and renewable energy dispatch will grow exponentially over the next decade. Energy tech companies (such as digital divisions within Siemens Energy and Schneider Electric, as well as numerous smart grid startups) consistently report that the hardest candidates to find are exactly those who understand both the operational logic of power systems and modern ML engineering paradigms — which is precisely this Brazilian engineer's potential competitive edge. It's also worth noting that Latin America's homegrown energy tech ecosystem is maturing rapidly. Brazil, Chile, and Colombia have all produced startups focused on smart grids and renewable energy AI, and these companies have a strong preference for engineers familiar with local regulatory frameworks and grid characteristics — giving this engineer an additional moat against international candidates.

If this engineer combines his deep understanding of the power sector with AI capabilities, he's not entering a red ocean — he's looking at a relatively blue-ocean vertical niche.

Closing Thoughts: This Is More Than a Job Change

The value of this story lies in how it goes beyond the surface question of "should I switch careers" to touch on something more fundamental about how technical professionals make career choices: What are we actually pursuing? Salary and stability, or freedom, growth, and mental wellbeing.

For this engineer, the answer may not be a binary choice between "electrical engineer" and "AI engineer" — but rather finding a path that brings together his existing expertise, industry advantages, and the lifestyle he actually wants. The skills are already there. What remains is courage and strategy.