This video features an interview with Janvi Kalra, a software engineer who transitioned into AI engineering. The conversation covers her career path, including internships at Google and Microsoft, her experiences at startups like Coda and OpenAI, and her insights into the AI job market and the skills needed for success in the field.
This video features an interview with Janvi Kalra, a software engineer who transitioned into AI engineering. The conversation covers her career path, including internships at Google and Microsoft, her experiences at startups like Coda and OpenAI, and her insights into the AI job market and the skills needed for success in the field.
What specific projects did Janvi work on during her internships at Google and Microsoft? The transcript states she worked on the search team at Google, exploring their internal tools and documentation. At Microsoft, she was on the Azure OS team, working on a product similar to Dropbox for Azure blobs.
What criteria did Janvi use to evaluate the "large market" aspect of potential startups? The transcript doesn't detail Janvi's specific criteria for evaluating market size, only mentioning it as one of four factors (along with revenue, customer loyalty, and competition) in her rubric for assessing startups.
What were some examples of the "unknown harms" Janvi mentioned in relation to increasingly powerful AI models? Janvi mentions the challenge of identifying "unknown harms" that might arise from more powerful models and the difficulty of anticipating how users might jailbreak or exploit them. She doesn't provide specific examples from the transcript.
What specific resources (books, online courses, etc.) did Janvi find helpful in developing her AI engineering skills beyond hackathons and self-learning? Janvi mentions using resources like "Cracking the Coding Interview," NeetCode's videos for LeetCode prep, and Alex Xu's books for system design. She also mentions Langchain's documentation and notes the book by Chip Huyen on AI engineering. She emphasizes learning by doing through hackathons as the most effective method.
Ask anything...
Regarding a specific curriculum: Janvi doesn't explicitly mention following a structured curriculum for learning AI engineering. Her approach was largely self-directed, driven by practical projects and addressing real-world challenges through hackathons and her work experience.