This was my first role where I served as the Product Manager (or a similar-role), and the start of my interest in PM. Given that this was for a Series-A green-tech startup, I was able to get a lot of firsthand responsibility, so I was able to learn a lot.

One of the biggest lessons from this project was how central data quality and system constraints are to building a product that is actually usable in the real world. Early on, it was tempting to design idealized dashboards and recommendation views based purely on what would be most helpful for government officials. However, as we moved closer to implementation, it became clear that the reliability, freshness, and consistency of underlying data would ultimately shape what the product could responsibly promise.
Working with engineers forced me to confront questions I hadn’t initially foregrounded: How frequently does energy consumption data update across different buildings? What happens when data is missing, delayed, or reported in inconsistent formats? How do we distinguish between true anomalies and normal seasonal variation? These considerations directly influenced interface decisions, from how trends were visualized to how confidently the product could recommend upgrades or interventions.

From a product perspective, this experience taught me the importance of treating engineers as early thought partners, not downstream implementers. Some of the most productive moments in this project came from pressure-testing ideas together—walking through how a feature would actually be supported by backend pipelines, what assumptions it relied on, and where the system might break. In several cases, this led us to simplify or reframe features so they aligned better with what the data could reliably support, rather than over-indexing on surface-level polish.

If I were revisiting this project today, I would invest earlier in defining data schemas, confidence thresholds, and update cadence alongside the core product requirements. Doing so would allow product decisions to be grounded in technical reality from the outset, rather than adjusted retroactively. More broadly, this project reinforced for me that good product design isn’t just about clarity at the interface level—it’s about building trust between the system and its users, which ultimately depends on thoughtful coordination between design, data, and engineering constraints.