Best Practices for Implementing Serverless Functions in a Microservices Architecture

[Onam]

I'm building a microservices architecture using serverless functions, and I want to ensure I follow best practices for scalability, reliability, and cost optimization. What are the key considerations and best practices for implementing serverless functions in a microservices setup? Any tips for managing dependencies, orchestration, and security in this context?