Brands receive customer feedback from many sources simultaneously: review platforms like Trustpilot and Google, social media like Twitter and Reddit, app stores, press articles, and forums. The problem is that this information is scattered, unstructured, and voluminous. A company might have 500 reviews on Trustpilot, 200 Reddit mentions, dozens of press articles, and hundreds of tweets per month. Nobody has time to read all of it.So what happens? Teams either ignore most of it, or they sample randomly, or they wait for quarterly reports that arrive too late to act on. Meanwhile, patterns emerge that nobody sees: the same product complaint showing up across three platforms, a sentiment shift after a campaign launch, a specific store location accumulating bad reviews.And when an executive asks "how's our brand doing?" before a board meeting, someone scrambles to pull data manually, read through reviews, and write a summary. It takes hours or days. Often the request comes with a 2-hour deadline.Existing tools help with collection but not comprehension. They aggregate mentions and count stars, but they don't understand what customers are actually saying. They can't distinguish relevant mentions from noise, especially for brands with common names.We used generative AI to read and analyze every piece of feedback across all sources. It classifies sentiment, extracts recurring themes, detects anomalies, and generates analytical reports on demand. Need a report for the board in 10 minutes? It's there. The AI understands business context, filters out irrelevant mentions, and focuses on what matters.The problem we solve is the gap between having customer feedback everywhere and actually understanding what customers think, when you need it.

The project was concluded very recently, so we're measuring early results from the original client rather than large-scale statistics. Here's what we observed:Missed opportunity recovered. During the first week of use, the client discovered a highly positive review from an influencer they hadn't noticed. That review led to a conversation, then a business partnership. Without Pulse surfacing it, the mention would have been buried among hundreds of others across platforms. One caught signal, one real business outcome.Report production time reduced from one week to five minutes. Before Pulse, preparing a brand health report for stakeholders meant manually collecting reviews from multiple platforms, reading through them, identifying patterns, and writing a summary. This typically took a marketing team several days to a full week depending on volume. With AI-generated reports, the same deliverable is ready in minutes. The team reviews and adjusts if needed, but the heavy lifting is done.Product launch monitoring without the overhead. Marketing teams used to assign someone to manually track customer reactions after a launch : scrolling through Twitter, checking Reddit, reading new reviews daily. It was tedious and inconsistent. With Pulse, they get a live dashboard showing sentiment evolution and emerging themes without dedicating headcount to manual monitoring. The time saved is reallocated to actually responding to feedback rather than hunting for it.Faster anomaly detection. Negative spikes that previously went unnoticed for days or weeks now trigger alerts within the collection cycle. Early detection means earlier response, which limits reputation damage.

This project exists because a client asked for it.They came to us with a problem: customer feedback everywhere, no way to process it at scale, and a team spending hours manually tracking what people said about them online. They couldn't afford a custom internal tool. They told me : build this as a product, we'll be your first user, and make it available to everyone facing the same problem.So we built it together. Their real needs shaped the features. Their feedback refined the AI. Their daily usage proved what worked and what didn't. Pulse isn't a solution designed in isolation, it was co-created with the people who needed it most.That origin story matters for this award:Problem-first, not technology-first. We didn't start with "let's use Gen AI for something." We started with a genuine business problem and discovered that Gen AI was the only way to solve it properly. Reading and understanding thousands of unstructured mentions across platforms is exactly what large language models enable.Built with users, not for users. The first client wasn't a beta tester receiving a finished product. They were a partner in the design process. Every feature exists because it solved something real.AI as foundation. Pulse couldn't work without generative AI. Sentiment analysis, theme extraction, contextual filtering, report generation : the entire product relies on AI comprehension. It's not an add-on. It's the core.Already delivering value. This isn't a prototype. Real clients use it daily. Real results: a partnership discovered from a surfaced review, reports generated in minutes instead of days, product launches monitored without manual overhead.We built something useful, with the people who needed it, powered entirely by Gen A. https://pulse.kalaia.fr/login