Tejus C.

Another Classic case for Financial Engineering  1. Business Value Measure  - If the new feature provides significant business value or competitive advantage, it should be worth prioritizing over addressing tech debt, even it leads to longer term challenges.  2. Maintainability Measure - The ease with which new engineers can understand, maintain, and innovate within the system directly impacts feature delivery.A low maintainability measure means even if you prioritize new features one cannot deliver on time. This should definently be a data driven measure of the above metrics. Even if one says customer is the boss someone should pay for the extra effort we are putting in so we'll have to financial engineer our way out.

Don't ask a non machine learning entrepreneur for advice.

Unpopular Opinion but true. 1. CS curriculum in undergrad at most universities is super solid to get your fundamentals right. 2. Study your subjects well, understand the history and the current research that's being done in those subjects. 3. Build foundational software like operating systems, compilers, databases. 4. Write assembly code, implement your own caching mechanisms in your microcontroller and learn to verify the correctness of the algorithms and programs you write. Like Dijkstra said "Computer science is no more about computers than astronomy is about telescope".

Hackers and Methodologists approach. 1. Initially the goal has to be to get to the solution by whatever means, Once its proved it can be done by certain ways then you find a middle ground to make the approach systematic and scalable. 2. Hackers can contribute to get to the solution quickly. 3. Then you bring in your Systematic thinkers to collaborate and understand the process to get to the solution.

Diversify. 1. Learn a language like C++/Rust to navigate codebases and implement foundational system softwares like databases. 2. Learn a language like Java to build enterprise applications where transaction support is very important. 3. Learn a scripting language like python/javascript for fast POCing/working with machine learning frameworks and web frameworks.

Stay close to academia. Many times, breakthroughs in apparently unrelated domains lead to innovation in computer science. The idea of superposition, which captures the dual particle and wave aspect of everything, was transformed by the Schrodinger's cat experiment. This led to advancements in semiconductors, which are fundamental to the existence of computers.