Latest news with #Qualitara
Forbes
26-06-2025
- Business
- Forbes
The Future Of Engineering Is Human-Led And AI-Powered
Sebastian Avila is the co-founder at Qualitara. In the past two years, AI tools have become deeply embedded in the way engineering teams operate. From generating boilerplate code to assisting with architecture diagrams, AI is now part of the daily workflow. But despite this rapid adoption, the demand for experienced engineers continues to rise. This shift is creating a new dynamic in modern engineering teams, where success depends on how well human expertise and machine capabilities work together. From Writing Code To Guiding Systems Tools like Cursor and Windsurf are eliminating repetitive coding tasks. Engineers are spending less time on boilerplate and more time reviewing, integrating and validating AI-generated code. The role of a senior engineer is becoming less about doing every task and more about making sure the right decisions are being made across the system. They are setting direction, validating outcomes and ensuring that the software being shipped actually serves the business. Like autopilot still requires an experienced pilot for critical decisions, engineering teams still need experienced leaders who know when to intervene, how to adapt and where to go next. Where AI Performs Well AI delivers real value in several areas: • Identifying bugs, inconsistencies and code smells across large codebases. • Generating documentation and summarizing legacy code with speed and accuracy. • Enforcing standardized patterns across microservices and infrastructure. • Automating unit tests, regression tests and common implementation details. • Accelerating code generation by helping developers move faster on standard components and avoid 'blank screen' delays. • Assisting in code reviews by flagging common issues and suggesting improvements early in the development cycle. These capabilities improve consistency and free up time that engineers used to spend on repetitive tasks. Where AI Falls Short And Human Judgment Takes Over Despite these advances, AI tools are still limited in ways that become clearly evident without strong senior oversight. They are only as good as the data they have seen and the patterns they can generalize. That is not enough when it comes to solving nuanced, evolving or high-stakes engineering problems. Here are key areas where AI continues to struggle and where senior engineers play a critical role: AI tools can suggest patterns or optimize resource usage, but they lack the foresight and trade-off analysis required in architectural decisions. They cannot anticipate how a system needs to scale over time, deal with edge-case reliability concerns or weigh performance against maintainability. In enterprise settings, architect-led reviews typically uncover issues in AI-generated proposals. Often, these adjustments involve rethinking core components because the AI lacked an understanding of business context, security policies or long-term goals. Forrester emphasizes ongoing human validation as essential. Only experienced engineers can make these calls effectively. They factor in company-specific constraints, regulatory risks and organizational readiness, which AI is not equipped to evaluate. AI lacks an inherent understanding of secure-by-design principles. It can unintentionally replicate insecure code patterns from public datasets, omit essential validation steps or suggest outdated dependencies with known vulnerabilities. For example, an AI tool once suggested downgrading a runtime environment to fix a compatibility issue. However, a senior engineer recognized that the recommended version was no longer supported and posed security vulnerabilities. Their intervention prevented a potentially critical exposure. Senior engineers play a crucial role in reviewing AI outputs through a security lens. Their knowledge of up-to-date practices, threat models and system dependencies helps ensure that AI-assisted development does not introduce avoidable risks. AI tools can generate a wide array of test cases based on past failures or code coverage gaps. However, they frequently miss critical business-specific scenarios or edge conditions that are not well represented in historical data. Senior quality assurance (QA) leads often catch the majority of high-priority test gaps. They understand what absolutely must not fail and apply domain-specific knowledge that AI cannot learn on its own. When AI-generated fixes introduce regressions, it is experienced engineers who step in to restore system stability. Why Senior Engineers Matter More Than Ever AI is making development faster but not necessarily better unless experienced engineers are there to guide the process. Their value is not in doing what AI can do but in knowing what AI cannot. Senior engineers today are doing more of the following: • Evaluating when AI-generated solutions are viable and when they need to be challenged. • Balancing immediate gains with long-term architecture integrity. • Translating between business needs, user behavior and technical feasibility. • Designing resilient systems that evolve with the product, not just meet today's needs. • Coaching developers to use AI productively and with critical thinking. These functions cannot be automated. In fact, as AI becomes more deeply integrated, these roles become even more vital. Teams that lack senior oversight often find themselves drowning in technical debt, security gaps and disconnected solutions that fail to meet real needs. Data supports this shift. Research from IBM and MITSloan indicates that engineers trained or mentored in AI tool usage can see productivity gains of up to 40% or more. But without mentorship, adoption plateaus and quality suffer. The best-performing teams are those where human and machine strengths are combined, not where AI is left to run unsupervised. Additionally, while AI enhances productivity, organizations must recognize that senior oversight remains essential. Continued investment in experienced talent is critical to maintaining quality and managing the increased complexity that comes with integrating AI. AI Is Not A Replacement—It Is An Accelerator The evidence is clear: AI is not here to replace senior engineers. It is here to support them, to extend their reach and to elevate the quality and speed of the work they deliver. Far from making human expertise obsolete, AI is increasing the value of experience, judgment and leadership in software development. The most successful engineering teams are not those that automate everything but those that learn how to combine human insight with machine efficiency. Senior engineers who embrace this shift are not becoming less relevant—they are becoming more strategic, more impactful and more essential than ever. AI is not making great engineers disappear. It is making them even better. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Forbes
28-03-2025
- Business
- Forbes
The Hidden Cost Of Bad Software Practices
Sebastian Avila, co-founder at Qualitara. getty Software isn't just a tool; it's the backbone of modern business. Yet, poor software practices silently drain billions of dollars from organizations every year, crippling innovation, inflating budgets and derailing projects. The numbers speak for themselves: • $2.41 trillion was the estimated cost of poor software quality in the U.S. alone. • In poorly executed projects, 50% of software development budgets are wasted on bug fixes instead of delivering business value. • Late-stage defect detection can be 100 times more expensive than catching bugs early in development. • 70% of digital transformation initiatives fail, often attributed to mismanaged software execution and quality issues. When software fails, the consequences extend far beyond lost revenue. A single undetected error can trigger product recalls, security breaches and irreversible reputational damage. The Samsung Galaxy Note 7 recall is a prime example: late-stage defects in its battery management system caused devices to overheat and catch fire, forcing Samsung to recall millions of units. The recall and production halts resulted in a $17 billion loss. The lesson is clear: Bad software is expensive, and the later you catch defects, the higher the cost. But what's the solution? Fixing software problems starts before a single line of code is written—it starts with hiring the right people. Poor hiring decisions don't just impact payroll; they derail projects, slow innovation and inflate long-term costs. • A bad hire can cost as much as 30% of that employee's first-year salary. • An employee who underperforms takes 70% more time to manage than a high-performing one. • 60% of bad hires will negatively affect the performance of other team members. This is why top organizations care so much about hiring A-players—the top 10% of engineers—who don't just write code but solve problems before they escalate. They proactively identify risks, build scalable solutions and ensure that software is reliable from the start. To consistently hire A-players, companies should adopt a structured, repeatable approach to identifying top talent. A key element of this process is creating clear scorecards that go beyond standard job descriptions, defining the role's mission, key outcomes and required competencies. This ensures alignment with business objectives and team dynamics, while also enabling consistent evaluations across candidates, especially useful when multiple interviewers are involved, helping to compare apples to apples. Structured interviews can then be used to assess candidates based on real career experiences rather than hypothetical scenarios. By exploring past roles, achievements and challenges, companies can uncover patterns of success, adaptability and problem-solving skills. Finally, rigorous reference checks provide an additional layer of validation. Instead of generic inquiries, they should focus on performance patterns and insights from former managers and colleagues. Cross-referencing a candidate's statements with past supervisors' perspectives can highlight consistency and credibility. Beyond the standard questions, it's essential to dig deeper into how the candidate responded to feedback, influenced team dynamics and handled setbacks. Asking for specific examples of their problem-solving approach, how they navigated conflicts and what their manager might have changed about their performance can reveal crucial insights. This structured hiring approach, inspired by principles from Who: The A Method for Hiring and insights from The Manager's Handbook, enhances consistency and increases the likelihood of securing high-performing talent. However, hiring great people isn't enough. Without strong engineering standards, even the best engineers can't deliver consistent, high-quality results. Many companies mistakenly believe that hiring top-tier engineers automatically leads to high-quality software, but even the best engineers can't thrive in a chaotic environment. Talent without the right guardrails leads to inconsistency, while guardrails without talent lead to stagnation. To create scalable, high-quality software, organizations must establish clear engineering standards that ensure everyone follows a cohesive approach. While the specific methodologies will vary between teams, companies should aim to implement industry-proven best practices that help drive reliability and efficiency. Some examples include: • Shift-Left Testing: Catch defects early by prioritizing testing in the design and development phases, reducing late-stage rework. • Continuous Integration/Continuous Deployment (CI/CD): Automate testing and deployments to improve release velocity while maintaining stability. • Automated Testing: Use tools like Selenium, Jest or Cypress to detect issues before they reach production. • Static Code Analysis: Tools like SonarQube help spot vulnerabilities and anti-patterns before they become production problems. • Security Best Practices (OWASP): Enforce secure coding standards to prevent costly security breaches. Organizations can also enhance their software development efficiency by adopting proven frameworks like DORA (DevOps Research and Assessment). DORA provides a data-driven approach to measuring and improving engineering performance by benchmarking teams against elite engineering organizations. It focuses on four key metrics that directly impact software delivery and operational efficiency: • Deployment frequency • Lead time for changes • Change failure rate • Mean time to recovery (MTTR) By tracking these metrics, companies can identify bottlenecks, optimize workflows and measure progress against industry-leading teams. While DORA is a widely adopted framework, it is just one of many approaches that organizations can use to continuously refine their engineering processes and drive long-term results. Success comes from cultivating a measurable environment where talent and best practices align with business goals. Building high-quality software takes both talent and strong standards. One without the other simply isn't enough. Organizations must invest in both: • A-players with strong technical skills who understand the business impact of their work. They proactively drive quality, communicate clearly and consistently raise the bar by delivering meaningful outcomes—not just completing tasks. • Engineering best practices designed to ensure consistent performance, eliminate inefficiencies and align with industry benchmarks. This balance separates high-performing engineering teams from those stuck in a cycle of technical debt, rework and stagnation. Companies that get this right don't just build better software—they save time, reduce operating costs and improve productivity. Ultimately, engineering excellence isn't just about writing code—it's about building a system where top talent and best-in-class processes consistently deliver exceptional software in a predictable, repeatable manner. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
