AI Augments Software Engineering--Focus on Problem-Solving, Not Code - Episode Hero Image

AI Augments Software Engineering--Focus on Problem-Solving, Not Code

Original Title:

TL;DR

  • AI-driven code generation risks creating billions of lines of ungoverned technical debt, necessitating platforms that integrate AI with existing enterprise assets for governance and stability.
  • The "80% that matters" in software engineering involves planning, system architecture, and cross-team coordination, areas AI currently struggles to automate, leaving these skills crucial for career longevity.
  • Agentic workflows, by automating back-office tasks, enable human employees to focus on customer engagement and revenue-driving activities, leading to significant business growth like 20% increases.
  • Low-code platforms are evolving to "agentic low-code," blending AI's probabilistic reasoning with deterministic enterprise systems to deliver innovation at pace while maintaining governance and security.
  • The future of software engineering lies in leveraging abstractions and the right tools for the job, rather than specializing in a single technology, fostering adaptability and broad skill development.
  • Companies are shifting team structures and roles to embrace rapid iteration and AI integration, streamlining processes by getting engineers closer to customers and removing unnecessary steps.
  • The primary value of AI in the enterprise will be in implementation and adaptation to specific business problems, driving revenue or cost savings, rather than purely technological advancements.

Deep Dive

AI is not poised to replace software engineers by automating code generation, but rather by rendering irrelevant those who focus solely on coding. The true value and future of software engineering lie in the remaining 80% of the job--problem-solving, architecture, integration, and customer engagement--an area where AI currently offers limited capabilities. This shift necessitates a reorientation of skills from mere code production to strategic application of technology for business impact, creating significant opportunities in implementation and adaptation.

The rise of agentic workflows, powered by AI, is transforming how businesses operate by automating complex, time-consuming tasks. For instance, Travel Essence has seen a 20% growth by redeploying travel planners from back-office work to customer-facing sales roles, directly impacting bookings. Similarly, Graham Housing Finance in India uses AI agents to process mortgage applications, reducing human intervention to a mere 2% of cases and accelerating approvals for more buyers. These agentic systems are not limited to customer-facing roles; Access Bank uses them for internal log processing and performance optimization, indicating a broad applicability across operational and technical domains.

While AI excels at generating code, it struggles with existing enterprise assets and governance, creating a significant risk of "ungoverned code" if not managed through robust platforms. This is where low-code platforms like OutSystems play a crucial role. They provide the necessary infrastructure, tooling, and deterministic systems to integrate AI's non-deterministic capabilities effectively. The future of software engineering will involve a blend of high-code and low-code approaches, with AI acting as a tool to leverage existing enterprise assets and workflows rather than building from scratch. This approach allows for rapid innovation and development pace while maintaining the security, governance, and stability required for enterprise systems.

The evolving landscape demands a shift in career focus for software engineers. Instead of specializing in a single technology, the most successful individuals will be those who master the "80% of engineering work" that AI cannot replace. This includes understanding business problems, customer needs, and architectural design, and then leveraging AI and platforms like OutSystems to implement solutions. The key takeaway is to focus on applied AI--solving real business problems that generate revenue or reduce costs--rather than solely on the technology itself. This applied approach, particularly in implementation and adaptation, will be the primary driver of value and career longevity in the coming years.

Action Items

  • Audit 80% of job: Identify 3-5 non-code tasks (planning, architecture, integration) that AI cannot replace and focus skill development there.
  • Create problem-solver persona: Shift focus from code generation to understanding business needs and driving revenue through technology solutions.
  • Implement agentic workflows: Pilot agentic systems on 2-3 core business processes to automate grunt work and free up human capacity for customer engagement.
  • Measure impact of AI adoption: Track 3-5 key business metrics (e.g., customer acquisition cost, time-to-market) to quantify value derived from AI and agentic systems.
  • Develop hybrid skill set: Integrate low-code and high-code proficiency to leverage the right tool for specific business problems and enterprise assets.

Key Quotes

"The 80% of engineering work that AI cannot replace"

Woodson Martin explains that while AI is advancing rapidly, it cannot fully replace human engineers. He suggests that the core value of an engineer lies beyond just writing code, implying that other skills are more resilient to automation. This highlights a shift in the perceived essential skills for software engineers.


"The engineers who thrive are the ones who master the other '80% that matters.'"

Woodson Martin argues that success in software engineering is not solely dependent on coding ability. He emphasizes that mastering the remaining 80% of the job, which likely includes problem-solving, communication, and business acumen, is crucial for career longevity and advancement. This points to a broader definition of engineering expertise.


"Every new wave of technology that comes along adds complexity doesn't take it away."

Woodson Martin posits that technological advancements, including AI, tend to increase, rather than decrease, the overall complexity of enterprise systems. This perspective challenges the notion that new technologies will simplify existing IT landscapes, suggesting that managing this complexity will remain a key challenge.


"The best software engineers I know the most senior most experienced look at the artifacts they get from cloud code or any other ai and they're like you know the code kind of sucks like what I'm doing is I'm building tech debt at a pace you know 100 times faster than I could build tech debt before."

Woodson Martin shares a critical observation from experienced engineers regarding AI-generated code. He notes that while AI can produce code rapidly, its quality is often subpar, leading to the accumulation of technical debt. This suggests that human oversight and expertise are still necessary for maintaining high-quality, sustainable software.


"The challenge of managing enterprise systems in the future won't be the same as it was yesterday but it also isn't going away."

Woodson Martin asserts that the complexities inherent in managing enterprise systems will persist, even with the advent of new technologies like AI. He implies that the nature of these challenges will evolve, but the fundamental need for skilled management and architecture will remain. This underscores the enduring importance of foundational IT management principles.


"Get close to customers understand what people are really trying to solve there's so much hype today that's just purely about the tech and we're kind of then in our own little bubble talking about the magic of models and so much of what and there's a huge gap like I think everybody can see that between the trillions being invested in AI infrastructure... relative to the actual monetization."

Woodson Martin advises software engineers to focus on understanding customer needs and solving real business problems, rather than getting lost in technological hype. He points out a significant gap between investment in AI infrastructure and its actual monetization, suggesting that practical implementation and value generation are key to future success. This emphasizes the importance of applied AI and business impact.

Resources

External Resources

Books

  • "The 80/20 Rule" - Mentioned as a principle for understanding job responsibilities.

Articles & Papers

  • "AI Won't Replace Software Engineers, But This Might (CEO Perspective)" (Beyond Coding Podcast) - Mentioned as the episode title and topic.

People

  • Woodson Martin - CEO of Outsystems, featured guest discussing AI, low-code platforms, and software engineering careers.

Organizations & Institutions

  • Outsystems - Mentioned as a leading AI-powered low-code platform.
  • OpenAI - Mentioned in relation to their Agent Builder and their adoption of low-code platforms.
  • Access Bank - Mentioned as a customer using Outsystems for internal agentic systems.
  • Graham Housing Finance - Mentioned as a customer using Outsystems for mortgage application processing.
  • Travel Essence - Mentioned as a customer using Outsystems for automating back-office work and improving customer engagement.

Tools & Software

  • Agent Workbench - Mentioned as an Outsystems tool for building agents.
  • N8n - Mentioned as an open-source option for building agentic workflows.

Other Resources

  • Agentic Workflows - Discussed as a key concept in modern software development and business problem-solving.
  • Low-Code Platform - Discussed as a technology enabling rapid application development and integration with AI.
  • Forward Deployed Engineer Model - Discussed as a strategy for integrating engineering expertise with business problems and customer domains.
  • Cobol Apps - Mentioned as legacy systems that companies are looking to modernize.

---
Handpicked links, AI-assisted summaries. Human judgment, machine efficiency.
This content is a personally curated review and synopsis derived from the original podcast episode.