Envié una solicitud electrónica. Acudí a una entrevista en Speechify
Entrevista
When you're done, we'll review your results. If you pass, you'll have a technical interview with an engineer on our team, with whom you can also ask any technical questions that you might have. After that, our final round interview is with our CEO, Cliff.
Preguntas de entrevista [1]
Pregunta 1
Project Overview
The project's goal is to implement an LRU Cache, an SSML Parser, and a helper function for converting an SSML Node Tree to a plain text string.
Task Details
The questions to solve are in src/main/java/com/speechify and the tests can be found in src/test/java
Implementation Checklist
createLRUCache: Implement a LRU Cache Provider with get and set methods.
parseSSML: Implement a SSML Parser that takes a SSML string and returns a SSML Node Tree.
Do NOT use or refer to any pre-existing XML parsing libraries.
ssmlNodeToText: Implement a function that takes a SSML Node Tree and recursively converts it to a plain text string.
Time to Implement
1 Hour 30 Minutes
Development Guidelines
Do's
Write clean, maintainable, and well-documented code.
Please follow the best practices and coding standards.
Test cases are provided for all methods; use them to ensure that your code is correct and meets our requirements.
You are free to use any official documentation or language references (MDN, Node Docs, etc).
You can use the debugging tools and native IDE features (only standard Auto-Completion)
Don'ts
Do NOT use any external libraries for the implementation.
DO NOT use any Coding Assistants like GitHub Copilot, ChatGPT, etc or any other AI based tools.
DO NOT visit direct blogs or articles related to implementation of the tasks.
DO NOT use Stackoverflow or any other forum websites.
Weird and invading my privacy, was good in general but not smooth and didn’t feel good while doing it, they should respect candidates privacy, it’s very disrespectful interview , first project was screen recorded all the time even when entering my git password
Envié una solicitud electrónica. Acudí a una entrevista en Speechify en may 2026
Entrevista
I applied for the job and received an automated email one minute later with a technical challenge using AI.
They provided a large repository for download and asked me to keep my webcam and screen recording active for 90 minutes (probably for training AI models).
They didn't provide any tools. You had to use your own tokens for the tests for 90 minutes without knowing if anyone was going to see them. Basically, that's what happened to everyone who left a review here.
I understand and like pair programming with AI, but doing it this way is impersonal, probably the worst way, and zero privacy-oriented. You never get feedback. It's one of the worst things I've seen this year.
Applied online on WTJ. Not 5 minutes after submitting your application *on a Sunday*, you receive an email that informs you that you made it ro the first round! Which is an automated cosing assessment. No human in the loop. No review of your application. They just want to give 90 minutes that you can't claim back without prior even checking that your expectations really mat h the role. "If all goes well, you'll talk with our CTO", they say. Two red flags: what is he measure of success? Why is the CTO involved at all? Looks to me like a bottlenecked process thatsjust a reflection of how they operate. Not a company I want to work with (not even for).
Preguntas de entrevista [1]
Pregunta 1
Take a bkind coding assessment with your cam and mic onao that we can spy on you.
