The Pragmatic Engineer
By Gergely Orosz
Big Tech and startups, from the inside. Highly relevant for software engineers, AI engineers and engineering leaders, useful for those working in tech.
33 episodes
All Episodes
Prioritizing Software Fundamentals to Manage AI-Generated Codebase Entropy
AI agents increase software entropy, which turns messy codebases into a liability. You can master the system by applying classic engineering mental models to guide agent reasoning. This approach transforms your codebase into a high-performance, sustainable development environment.
View Episode Notes →
Shifting Engineering Value From Syntax To Intent Architecture
AI is turning standard coding tasks into commodities. The value of an engineer is moving away from manual debugging and toward high-level architecture. To succeed now, you must build systems that agents can navigate, while actively working to make your own infrastructure obsolete.
View Episode Notes →
Designing Software Architecture for Hardware Performance and Longevity
Premature optimization is a myth used to excuse poor design. When engineers align software architecture with hardware realities, they avoid costly rewrites and build systems that remain performant, flexible, and durable.
View Episode Notes →
Shifting Engineering Value From Syntax Mastery To Systemic Accountability
AI agents are unbundling software engineering, moving the focus from syntax mastery to systemic accountability. Engineers who curate and verify AI output instead of simply accepting it will define the next era.
View Episode Notes →
Mastering System Verification to Counter AI-Generated Reliability Debt
AI-driven code generation creates an illusion of productivity that hides a decline in system reliability. Master this transition by treating production as a development stage and building a competitive moat through verification.
View Episode Notes →
Leveraging Lightweight Formal Modeling for Reliable Distributed Systems
The fast pace of software engineering often masks a lack of formal rigor. You can gain a competitive edge by mastering lightweight formal modeling to verify complex systems that AI currently cannot reliably reason about or secure.
View Episode Notes →
Why Autonomous AI Coding Factories Create Long-Term Technical Debt
Automated systems that operate without oversight create a trap where autonomous AI agents quietly degrade the maintainability of your codebase. Instead of relying on high-risk factories run entirely by agents, use slow loops and human-led architectural planning. This approach leads to more sustainable software quality and consistent long-term velocity.
View Episode Notes →
AI Tools Shift Engineering Value Toward Systemic Problem Solving
AI automates routine coding, but it makes true engineering more difficult by shifting the value from raw output to deep reasoning. Master system architecture and foundational rigor to remain competitive as the definition of seniority changes.
View Episode Notes →
Prioritizing Social and Cognitive Discipline Over AI Output Speed
AI-generated code creates the appearance of progress while undermining system reliability. You can master software engineering by prioritizing cognitive social friction and rigorous verification over raw output speed. This approach builds systems that are both lasting and verifiable.
View Episode Notes →
High-Agency Decision Making as a Durable Engineering Advantage
AI-driven speed creates a dangerous skill gap for engineers. Build a durable career moat by mastering high-agency decision-making and systems thinking, so you can navigate complex trade-offs when automated tools fail.
View Episode Notes →
Prioritizing Practical Control Over Dogmatic GitOps Workflows
GitOps is a misleading term. It causes friction when teams try to force imperative tasks into declarative containers. Instead, focus on short feedback loops, managing feature flags, and rolling forward to build resilient, scalable systems.
View Episode Notes →
How Personal Agency Shapes Technological Impact
Kelsey Hightower didn’t climb the tech ladder--he rewired it. By aligning small, consistent bets on learning and visibility with seismic industry shifts, he turned technical work into lasting influence. Real impact isn’t shipped in code--it’s built in trust, visibility, and the courage to solve problems no one sees.
View Episode Notes →
AI's Paradox: Faster Code Obscures Engineering Challenges
AI accelerates code, but can obscure engineering challenges, leading to degraded quality and a false sense of progress. Build robust software by embracing friction and focusing on cohesion.
View Episode Notes →
Rust's Compiler-Driven Design Ensures Long-Term Software Reliability
Rust's "difficulty" is a feature, forcing patterns that yield extraordinary long-term reliability and security by preempting common programming pitfalls.
View Episode Notes →
Anders Hejlsberg's Career: Enduring Language Design and Developer Productivity
Enduring software tools require a holistic developer experience, not just code. Building them is a decades-long game focused on human productivity and integrated ecosystems.
View Episode Notes →
AI Agents Undermine Software Quality, Create Engineering Challenges
AI agents promise efficiency but subtly undermine software quality, creating more work. Understand these hidden consequences to retain a competitive edge in AI-driven development.
View Episode Notes →
Foundational System Design Decisions: Hidden Costs and Durable Advantages
Understand how subtle engineering choices create hidden costs and durable advantages, revealing the deeper dynamics that shape systems and the industry.
View Episode Notes →
AI Agents Amplify Judgment and Craft in Software Development
AI agents amplify senior engineers' judgment and craft, not diminish them. Master AI to unlock unprecedented productivity while maintaining essential human oversight.
View Episode Notes →
Necessity Forged Uber's Hyper-Growth Microservice Architecture
Necessity, not strategy, forged Uber's hyper-growth architecture. Radical innovation born from crisis created thousands of microservices and bespoke tools, yielding enduring competitive advantage.
View Episode Notes →
WhatsApp's Lean Blueprint: Ruthless Focus Built Global Giant
WhatsApp built a global giant by rejecting complexity and prioritizing ruthless focus, proving that deliberate simplicity and engineer trust create an unreplicable competitive advantage.
View Episode Notes →
AI Accelerates Software Engineering From Craft to Orchestration
AI is accelerating software engineering beyond manual coding to orchestrating agents, demanding a rapid redefinition of engineering skills and team structures.
View Episode Notes →
AI Rewrites Software Creation: From Autonomous Agents to Generalist Skills
AI is fundamentally rebuilding software creation, shifting engineering skills from deep expertise to rapid, iterative prototyping and broad adaptability.
View Episode Notes →
Mitchell Hashimoto Reimagines Software Development Architecture
AI redefines software development by automating code writing, review, and debugging, fundamentally changing how we build and deploy applications.
View Episode Notes →
Kotlin's Pragmatic Design and AI-Driven Intent Communication
Kotlin emerged from Java's stagnation to address developer pain points, achieving widespread adoption. Now, AI offers a new paradigm to communicate intent, potentially shrinking coding effort tenfold.
View Episode Notes →
AI Revolution Ushers Third Golden Age of Software Engineering
AI is not automating software engineering, but ushering in a new golden age by raising abstraction levels. Engineers will manage complex systems, not just write code.
View Episode Notes →
AI Agents Redefine Software Creation Through Architectural Vision
AI agents redefine software creation, shifting focus from code writing to architectural vision and prompt engineering. Developers now architect systems, directing AI collaborators to build at unprecedented speed.
View Episode Notes →
S3's Evolution: Strong Consistency, Durability, and Expanded Data Capabilities
S3 evolves beyond object storage, offering strong consistency without cost increases and introducing native support for structured data and AI embeddings.
View Episode Notes →
Innovation Thrives on Constraint, Not Complacency
Innovation thrives on desperation, not complacency. Discover how constraints drive true breakthroughs and the first-principles approach Oxide Computer uses to build modern hardware.
View Episode Notes →
Founding Engineer's Journey: From Startup Risks to AI Innovation
Discover how to thrive as a founding engineer, take on unglamorous tasks, and leverage AI to build autonomous websites that redefine the internet.
View Episode Notes →
Developers Own Code Security: Fix Flaws Before Exploitation
Developers must own code security, proactively preventing vulnerabilities beyond known exploits by rigorously validating all inputs and managing secrets, making security an integral part of creation.
View Episode Notes →
AI Shifts Software Engineering from Deterministic to Non-Deterministic
AI transforms software from deterministic to non-deterministic, demanding new engineering approaches and rigorous testing, while core human skills remain paramount for navigating this profound shift.
View Episode Notes →
Netflix Engineering: Scale, Studio Tech, and Pitch-to-Play
Netflix engineers build the future of entertainment, handling trillions of events daily with unparalleled autonomy and a culture of "unusual responsibility" driving innovation.
View Episode Notes →
Lattner: Compilers, Swift, and Democratizing AI
Chris Lattner reveals how LLVM, Swift, and Mojo democratize programming, from compilers to AI, by prioritizing modularity, accessibility, and developer control over hardware.
View Episode Notes →