Project Outcomes
Hometeam Ventures’ mission is to close the housing gap by investing in early-stage construction and housing innovators. Driven by this mission, they manage a growing database of thousands of companies, which involves manually receiving, categorizing, and tracking them. With this database growing every week, the process is time-consuming, laborious, and limits efficiency. This project aimed to develop an AI tool to automate the process, enabling faster tagging, tracking, and resulting in better investment decisions.
The first deliverable was an AI Tagging Engine that scans company data to spot missing tags and automatically fills them in with the help of a trained machine learning model. Stakeholders can review and adjust the results before updates are pushed back into an Affinity Database, keeping the process fast while ensuring accuracy. Alongside this is a Retrieval Augmented Generation (RAG) system, powered by GPT-4, which combines retrieval with reasoning to deliver context-aware answers about Hometeam Ventures and its portfolio. Instead of simply pulling data, it synthesizes insights and supports deeper knowledge discovery.
In addition, a scalable frontend interface was designed and built, bringing these machine learning and artificial intelligence systems together in one place. Built-in security safeguards ensure controlled access and protect data integrity, making the platform practical and secure for all users.
This solution streamlines the tagging process, making it faster, more efficient, and significantly more accurate, while ensuring seamless access to insights across the dataset. Hence, saving Hometeam Ventures time spent on tagging companies manually.
Key technology/tools used: Next.js, Figma, Python, GPT-4, Qdrant, Git, GitHub, Shell scripts
Visual output showing the frontend interface.
Attribution: Sergius Justus Nyah
No Comments.