HabileData

HabileData was hired by a property brokerage firm, Next Address, to gather, standardize, and validate large volumes of real estate data. The team of 6-10 employees collected data on thousands of property listings, including details such as property size, pricing, location, amenities, ownership history, and photographs. HabileData cleaned the datasets, removed duplicates, and formatted them into structured outputs compatible with the firm's CRM and website. The key deliverables were complete, verified property databases that allowed the firm's agents to offer clients more accurate and up-to-date listings.

The company was also contracted by AI Tinkerers, a company that uses AI-powered image recognition tools to track and reduce food waste. HabileData was tasked with annotating thousands of food images with precision. The project required annotation of more than 50,000 images of food waste, with HabileData creating bounding boxes, labeling food types, and categorizing waste portions according to predefined parameters. Weekly batches of annotated images were delivered, along with quality audit reports. The final deliverable was a highly accurate, standardized dataset aligned with the company's machine learning needs.

HabileData worked with MobiSystems, Inc., a company that provides real-time news aggregation and analysis, to accurately annotate and categorize large volumes of online news articles. The team processed thousands of online news articles each week, tagging entities, classifying news topics, and marking sentiments. They delivered weekly batches of validated datasets with an accuracy benchmark of over 98%. The key deliverables were high-quality annotated text files, progress reports, and a scalable workflow that could be fed directly into the client's AI pipeline.