Presenters
Source
Unleashing Unprecedented Engineering Velocity: How AI is Revolutionizing Product Development 🚀
It’s an exciting time to be in tech! We’re witnessing a dramatic shift in how product teams operate, largely fueled by the incredible advancements in AI-assisted coding. This isn’t just about writing code faster; it’s about a fundamental transformation in engineering speed, confidence, and collaboration. Over the past six months, Business Insider has experienced this firsthand, and the results are nothing short of remarkable.
The Great Reversal: Engineering No Longer the Bottleneck 🤯
For years, the prevailing wisdom in product and technology teams has been that engineering is the primary bottleneck. Product teams always have a long wish list of features, and the pace of delivery was dictated by the size and velocity of the engineering team. However, at the start of this year, something surprising happened: this dynamic flipped entirely.
We started hearing astonishing feedback:
- Beating Estimates by Astonishing Factors: Projects estimated to take 3-4 months were being completed in mere weeks.
- Engineering Managers Seeking More Work: Teams were burning through their allocated tasks faster than they could be scheduled, proactively asking for more.
- Preference for Solo Work: Some engineers, typically collaborative, found the setup time for team projects to be a drain on their delivery speed, preferring to work independently.
Initially, the leadership team suspected something was wrong, perhaps a consequence of a recent reorganization. But the data told a different story: we were shipping at record rates. There were no upticks in incidents or defects; things were, by all accounts, improving.
To quantify this shift, consider this: In 2024, the PR team issued 16 press releases, with only one focused on product. For 2025, just through mid-year, they’ve already published 36 product-focused releases. If this trend holds, we could see over a 20x increase in publicly shared product launches by year-end! This unprecedented speed was undeniable, yet the feedback from the team remained curious.
The AI Catalyst: Beyond Just Faster Coding 💡
Digging deeper, we identified the root cause: AI-assisted coding. While our engineering teams had experimented with these technologies for years, something clicked at the beginning of this year. Tools like Claude Code, Codex, and Copilot reached a level of maturity that led to mass adoption, transitioning from experimental tools to integral parts of the daily workflow.
The obvious benefit is the speed at which individual engineers can write code. But this is just the tip of the iceberg. The more profound impacts are on the psychology of engineers and their overall productivity:
- Boosted Engineering Confidence: AI acts as a non-judgmental, always-available partner, helping engineers produce idiomatic code. This significantly reduces the time spent researching best practices and debugging, freeing them to explore new fields with greater confidence and speed.
- Increased Risk Tolerance: With increased velocity and confidence, engineers are more comfortable taking chances. This includes incurring mild technical debt, knowing they can address it later efficiently. This agility allows for faster progress than ever before.
- Reduced Barrier to Entry: AI tools have lowered the barrier to entry for engineers working with unfamiliar codebases. This enables broader codebase ownership and cross-functional collaboration, a challenge previously difficult to overcome.
Organizational Ripple Effects: Reshaping How We Build 🌐
These engineering-centric changes have triggered significant organizational shifts:
- Growing Bias Against Replatforming: Replatforming initiatives are notoriously time-consuming, error-prone, and often yield minimal returns. The rapid evolution of AI-assisted development makes long-term platform commitments (6-12 months) increasingly risky, as building internally might soon become virtually cost-free. The buy versus build logic has been fundamentally scrambled.
- New Attitudes Towards Technical Debt: Technical debt, once a dreaded concept, is now viewed with a more pragmatic lens. In the pursuit of rapid delivery, engineers are more willing to make compromises, confident that refactoring will be a less costly affair due to AI’s efficiency.
- Comfort with Expendable Code: In a world of tight engineering resources, the idea of writing code that might not make it to production seemed insane. However, with AI accelerating development, some code is now seen as expendable. This allows for innovative approaches like multiple engineers solving the same problem simultaneously or engaging in work before securing full stakeholder buy-in.
Evolving Product Function: From Gatekeeping to Collaboration 🤝
Our exploration began with engineering outpacing product, sometimes leading to product becoming the bottleneck. We examined our product function to identify areas for streamlining and faster integration with engineering:
- Discouraging Gatekeeping: Traditionally, Product Managers (PMs) guarded engineers’ time. With accelerated engineering velocity, this “gatekeeping” is less necessary and can create tension. We’ve begun discouraging this practice.
- Slimming Down Requirements Gathering: High documentation bars can slow down the product process. We’re focusing on leaner requirements to enable earlier engineer involvement.
- The Relay Race to Basketball Game Evolution: We’ve shifted from a “relay race” model (product hands off a finished product to engineering) to a “basketball game” model, emphasizing early and continuous collaboration between product and engineering.
- Rethinking Prioritization: When capacity is no longer the primary constraint, the focus shifts. The existential question becomes: is this a good idea? This lowers the bar for initial exploration, allowing for more initiatives to be considered.
- Combating Product Bloat: Rapid shipping can lead to feature bloat and negative user experiences. Careful consideration is needed to avoid muddying interfaces.
- Assessing Organizational Readiness: It’s crucial to ensure the organization is well-positioned to capitalize on new initiatives. Compelling ideas can be wasted if the organizational structure isn’t aligned.
- Openness to Sunsetting: With the ability to ship rapidly, we must also be prepared to sunset products that aren’t performing or are competing with other priorities.
Tactics for Maximizing Speed and Collaboration 🛠️
To harness this unprecedented speed, we’re actively evolving our collaboration models:
- Early Engineering Involvement: Engineers are brought in as early as possible to collaborate, not wait for product.
- Embracing Visual Prototyping (Vibe Coding): Product managers are now creating functional prototypes using tools like Vibe Coding. This “show, don’t tell” culture reduces reliance on lengthy documentation and minimizes translation loss.
- Committing to Prototyping: We are now formally committing engineering teams to prototyping exercises and even projects without initial stakeholder buy-in, believing in the direction and the potential for buy-in.
- Bringing Hackathon Energy to Daily Work: We’re exploring ways to integrate the cross-functional, real-time requirement-setting and building aspects of hackathons into our regular workflows, especially for certain product types.
- Blurring Product and Engineering Roles: We’re empowering engineers to take on more product tasks (Product Engineering) and PMs to engage more with code (using Vibe Coding for commits or building small applications). This creates a positive synergy.
Your To-Do List for the AI-Accelerated Future ✅
If you’re looking to optimize your organization for this new era of speed, consider these actionable steps:
- Establish Meaningful Dialogue with Engineers: Deeply understand their relationship with AI-assisted coding. Uncover skepticism, true believers, and effective use cases.
- Clean Up Your Developer Experience: With increased engineering velocity, past minor issues in developer experience can become severe problems. Invest in a smooth onboarding and development process.
- Audit Your Product Function: Examine every aspect of the product function. Identify what’s working, what’s not, and what’s truly necessary. Think of it like a house of cards – each element supports another.
- Lean into Blurriness: Investigate the space between product and engineering. Explore opportunities for cross-disciplinary collaboration and learning.
- Explore Creative Collaboration Models: Look to hackathons for inspiration. How can you bring that real-time, agile approach into your daily operations?
- Prioritize Flexibility and Nimbleness: During restructuring, embrace flexibility, nimbleness, and reactiveness. Rigid structures will hinder your ability to capitalize on these new advantages.
We are at the very beginning of this journey. Things are evolving daily, and we’re continuously optimizing. While we’re already operating significantly faster, we believe we’re far from reaching the ceiling of our potential speed. The goal? To potentially 10x our team’s speed by this time next year. The future of engineering is here, and it’s incredibly fast and exciting! ✨