From Startup Founder to Staff Data Scientist at Meta | Chris Keating's Journey

Chris Keating had already carved out an impressive path in tech—working as a data scientist at Twitter and helping scale early-stage startups—before deciding to step away to explore founding his own company. But when he made the call to shut down that entrepreneurial venture, he found himself at a crossroads. Getting back into data science wouldn’t just be about applying to jobs. It meant retooling, leveling up, and filling in the technical knowledge he’d never formally been trained in.

What followed was a three-month sprint to rebuild not just his résumé, but his technical foundation. The end result? Offers from Meta, Instacart, and DoorDash, and ultimately, a staff-level role at Meta.

Despite years of experience in analytics and product roles, Chris knew he had gaps—particularly around statistics and causal inference.

He didn’t come from a traditional computer science or data science background. His academic training was in economics and psychology, and while he had learned a lot on the job, concepts like regression discontinuity or synthetic control had never come up in practice. Even when tools or experimentation platforms handled the analysis, he realized he didn’t fully understand the mechanics under the hood.

So when he committed to the job search in January, he also committed to a full rebuild of his technical depth. For two months, he treated the process like a full-time job, putting in 60-hour weeks, revisiting statistical theory, building case study frameworks, and learning how to think like an interviewer.

Chris didn’t rely on a single platform to guide him. He used different platforms for structured, topic-based breakdown of data science concepts, then went far beyond what the course offered. Every time he encountered a new term—whether it was regression discontinuity, Z-tests, or causal uplift modeling—he chased it down through YouTube videos, papers, blog posts, and case examples until he could both explain it and apply it.

One of the most effective strategies he developed was using ChatGPT to simulate mock interviews. He would feed in prompts and questions, talk through his answers aloud, then ask for critique. It became his personal coach, helping him stress test his thinking and spot blind spots he wouldn’t have otherwise noticed.

While he uses different platforms, Interview Query became a key resource for realism. He turned to it heavily for company-specific prep, especially as he entered final rounds with Meta, DoorDash, and Instacart. What stood out wasn’t just the depth of the content—it was the fidelity. The questions mirrored what actual data science teams were asking in real-world interviews, especially around business case framing and metric interpretation.

Chris had known about Interview Query since his time at Twitter, using it primarily for SQL practice. But during this job search, he started to see how the platform had evolved.

As he progressed deeper into interview loops, he upgraded to Interview Query Premium to access the full library of questions and interview guides. That’s when he began using it strategically—not just to study, but to replicate the flow and structure of actual interviews.

He appreciated that Interview Query wasn’t focused on generic algorithmic challenges. Instead, it asked questions that tested a candidate’s ability to reason through ambiguous problems, define metrics, and interpret business data.

Compared to LeetCode, which he felt leaned too heavily toward software engineering logic puzzles, Interview Query’s content felt purpose-built for data scientists. And more importantly, the company-specific sections gave him an edge. Even when the exact question didn’t show up in his interview, the tone and framing felt familiar—making it easier to stay composed under pressure.

Chris spent over $4,000 preparing for interviews—not including the hundreds of hours of study time. That investment included:

• A $2,000 recruiting coach, who helped with résumé building, scheduling, and offer negotiation
• $1,250 on six mock interviews via external platforms
• $250 on an external course
• $160 across two months of Interview Query and LeetCode
• A final mock interview

While the total cost was significant, Chris viewed it as an investment in his long-term earning potential and confidence. Each tool served a specific purpose—some helping with content, others with delivery. But in his view, the content and realism of Interview Query helped bridge the gap between preparation and execution.

One of the more unique twists Chris encountered during interviews was a SQL debugging task framed around AI. Instead of being asked to write a SQL query from scratch, he was presented with one supposedly generated by ChatGPT—and asked to identify what was wrong with it. That kind of real-world application felt like a signal of where interviews are headed: less about rote syntax, more about thinking like an analyst.

In other cases, especially at Instacart and DoorDash, the interviews became more conversational and layered. Chris recalled one scenario where a simple opening question led into a more complex causal inference discussion, probing not just his understanding of methods, but how he’d apply them and interpret results. These were the moments that validated the hours he’d spent digging deeper into each concept, well beyond what most prep materials provided.

Looking back, Chris credits his success to specialization and intentionality. He didn’t just study broadly—he studied with the goal of eliminating every weak spot. He sought out mock interviewers with hiring experience at his target companies. He paid for resume coaching from an ex-Amazon and Microsoft recruiter, not a generalist. And when it came to prep materials, he looked for realism and company specificity—two things Interview Query provided.

He also offered a suggestion to the Interview Query team: increase awareness of everything the platform offers. Having used it for years, he initially thought of Interview Query as just a place for SQL questions. He didn’t realize there were full-length courses and interview guides until he was deep into prep.

With that awareness, he says, he would have used it even more heavily.