Overview
This video is an in-depth interview with Michael Novati, a former senior staff engineer (IC7) at Facebook/Meta, who shares his experiences joining the company early on, insights into its engineering culture, his growth journey to becoming a top engineer, and thoughts on how emerging AI tools like large language models (LLMs) might impact software development.
Main Topics Covered
- Michael Novati’s early experience joining Facebook (Meta) as an intern
- The engineering culture and technical environment at Facebook pre- and post-IPO
- The IPO experience and its impact on employees
- Michael’s internal newsletter and openness within the company
- Working directly with Mark Zuckerberg and early engineering stories
- Types of engineers that impressed Michael and what defines a “coding machine”
- The potential impact of AI and LLMs on software engineering roles
- Michael’s approach to career growth, productivity, and working as an IC7
- Common traits among IC7+ engineers and advice on landing code faster
- Reasons for leaving Meta and reflections on talent, hard work, and luck
- Advice to his younger self and lessons on feedback and improvement
Key Takeaways & Insights
- Facebook’s early engineering culture empowered engineers heavily, fostering rapid innovation with tools and codebases built largely from scratch.
- The “move fast and break things” culture was about breaking norms and innovating quickly, not reckless coding.
- Post-IPO, the company matured with a stronger focus on stability and financial impact, changing engineering dynamics.
- Michael positioned himself as a “coding machine” by being highly productive, taking initiative on cross-org projects beyond his immediate team, and building strong credibility and trust.
- Taste and judgment—knowing which problems to solve and how to minimize impact when changing code—are critical skills that develop over time and separate top engineers.
- LLMs and AI tools are currently productivity enhancers rather than replacements but could fundamentally change the nature of coding in the future.
- High-performing IC7+ engineers share traits like extreme diligence, sharpness, conscientiousness, and high attention to detail.
- Rapid iteration with high-quality feedback is essential to grow as a software engineer.
- Luck (e.g., timing and company fit) and talent both play large roles in career trajectory, but hard work remains the most controllable factor.
- Receiving feedback as a way to improve rather than as judgment or approval is crucial for growth.
Actionable Strategies
- Start writing code early and often; don’t overthink before acting.
- Seek feedback from experienced and high-taste mentors, not peers at the same level.
- Actively incorporate feedback and iterate rapidly to compound improvement.
- Build credibility by minimizing bugs, communicating effectively, and understanding the broader impact of your code changes.
- Focus on problems where you have clear solutions in mind to maintain productivity, but gradually push into more ambitious projects to grow.
- Manage meetings carefully to protect deep work time; push back on unnecessary ones with manager support.
- Build relationships across your team and organization, especially with those responsible for code deployment and maintenance.
- Use AI tools and LLMs as productivity aids to speed up coding and routine tasks, while continuing to develop your judgment and domain knowledge.
- Reflect on your feedback as guidance to improve rather than as a pass/fail test to reduce pressure and grow faster.
Specific Details & Examples
- Michael joined Facebook in 2009 when there were about 200 engineers, primarily working in PHP (later evolved to Hack).
- He merged two internal task tools within a week early on without telling anyone, demonstrating initiative but learning about the importance of communication and impact.
- He once single-handedly removed thousands of legacy “preparable” classes from the codebase over several months.
- Michael worked with Mark Zuckerberg during a 2009 hackathon on the idea of emoji reactions to posts, which later became a standard feature.
- Facebook’s IPO in 2012 was a rational event internally, celebrated but grounded; stock vesting happened six months later, with many employees holding their shares long term.
- An internal newsletter Michael wrote sparked conversations but also friction with HR and executives, illustrating the risks and benefits of transparency.
- “Clown Town” was an internal humorous term for engineers who introduced silly bugs; Evan Priestley was a prolific engineer and role model.
- Michael described LLM adoption as akin to evolving from Vim to VS Code, with potential for even more transformative agentic AI workflows.
- He reported spending about 30% of his time on his team’s work and 70% on broader org-wide initiatives.
- An example of judgment was building trust with deployment teams so he could push code rapidly and confidently, even risking resignation if he caused a production failure.
Warnings & Common Mistakes
- Writing a lot of code without incorporating feedback or improving style and quality can frustrate reviewers and slow growth.
- Rushing big changes without considering other teams’ ongoing work or the impact on users can cause friction and bugs.
- Taking feedback as judgment or approval instead of constructive input can hinder learning and improvement.
- Overcommitting to meetings can destroy deep work time and reduce productivity.
- Being too rigid in prioritization (only working on tasks you immediately know how to solve) can limit career growth.
- Being openly critical or controversial internally can cause unintended political friction, even when motivated by transparency.
- Relying too heavily on stock price or compensation fluctuations rather than fit and performance can be a risky career approach.
Resources & Next Steps
- Michael mentioned internal tools at Meta like TBGS (code search), internal blogging via Notes, and continuous integration systems.
- He recommended seeking out senior engineers, skip-level managers, or widely respected people within your org for mentorship and guidance.
- For those interested in interview preparation and career development, Michael is involved with formation, a platform aimed at helping engineers improve and prepare.
- Viewers are encouraged to subscribe to the channel and follow Michael on LinkedIn, Reddit, or other platforms for follow-up discussions and advice.
- Embracing AI tools and continuously experimenting with prompts and workflows is suggested to stay ahead in productivity.
- Reflecting on feedback and adopting a growth mindset is emphasized as a personal development approach.