I'am currently doing a non-research Dipl.Eng (Engineering Diploma in French school system equivalent to MSc) in Computer Science.
The courses I am currently taking are software oriented, there's an absence of anything related to TCS except an Algorithm Design course I took in my first year. However my personal interests are in TCS,in the last two years (prior to starting the program I finished a BSc in Applied Mathematics) I did a deep dive in TCS and Cryptography my goal was to understand zero-knowledge proofs so I read many books (Prof. Avi Wigderson's Mathematics and Computation, the first part of the Computational Complexity book by Arora and Barak, Scott's Aaronson book + lecture notes). I also started reading papers released on eprint.iacr.org website about zkSNARKs and zkSTARKs and I implemented some of the works.
I would like to pursue a PhD in TCS later on (I am graduating in June 2022) but there are no programs available in my country, I was told from a professor in a foreign university that the field is very competitive and because I have no published results I have very slim chances of getting accepted.
Is there an absolute necessity to publish at my level, if so what kind of research should I work on ? Improve some SNARK scheme ? Make a synthesis of previous results ? Publish in an adjacent field where I have some results (Machine Learning) ?