I participated in Bayer’s Data Engineer hackathon in December 2025 and unfortunately had a very disappointing experience with the assessment process. The hackathon is structured in two sprints: Sprint 1 (group activity): Understand the problem, design architecture & data model. The entire group is evaluated and presented together. Sprint 2 (individual): Actual implementation in Databricks – but only for candidates who clear Sprint 1. I was eliminated after Sprint 1. The biggest issue is that Sprint 1 performance is judged at group level, yet Sprint 2 (the actual coding) is individual. This means if you land in a weaker or less vocal group, you pay the price even if you personally contributed good ideas. Several participants (including me) felt we weren’t given a real chance to showcase individual strength because the evaluators (Chapter Lead + Architect) seemed to have very specific preconceived solutions in mind. Any deviation or alternative approach—even if valid—was dismissed. The evaluators rarely probed individual contributions during the presentation; they judged the final deck and the group’s presentation skills. People who were quiet or got overshadowed (common in newly formed mixed-experience teams) were automatically filtered out without ever writing a single line of code or demonstrating hands-on Databricks skills—the core requirement for a Data Engineer role. They also infront of our team crossed the entire squad which was unprofessional as we should never demean the candidates and be extremely polite with them. It felt more like a test of “read the interviewer’s mind” and “get lucky with strong teammates” than a fair evaluation of data engineering capability. Ironically, the implementation round that actually matters is individual, but many capable candidates never reach it because of this flawed group filter. Pros: Real-world problem statement, good exposure to Databricks environment for those who make it to Sprint 2. Cons: Heavily team-dependent elimination in round 1, rigid expectation of one “correct” architecture, almost no opportunity to demonstrate individual technical depth. I would strongly advise future candidates to be very vocal in the group and try to align exactly with what the evaluators hint at, because individual brilliance won’t save you if the group output doesn’t match their exact mental model. I came out feeling the process was unfair and not reflective of real data engineering skills. Hope Bayer reconsiders making the architecture round individual or at least evaluates individual contributions properly. Would not recommend applying if the role uses this hackathon format.

Very pleasant and professional, only 1 round, technical questions regarding previous projects and how they relate to their operations. Got an offer but could not accept based on ongoing interviewing