Element AI is an artificial intelligence software company that helps enterprises operationalize AI. Element has several AI tools, applications and products that serve different industry verticals including insurance, banking, finance, retail and manufacturing.
Contributed to various projects and products across the Element AI including - Initiating early product design strategy for products, high-profile customer projects, investment decks, contribute to innovative AI research, manage, hire and grow the design team.
Here are two products I contributed product design strategy:
Document Intelligence
Automate field extraction from any document, boosting throughput and maximizing time spent on high-value tasks.
Collaborated with AI developers and scientists to research innovative solutions to improve document processing for clerks and end users.
Innovated work was done in order to determine the most optimum workflow to process documents quickly, correctly and accurately. AI models are only as good as the data it receives, as more quality data is available the models are able to begin moving from auto-fill to predicting outputs. Novel and dynamic user workflows needed to be considered for this product.
Trade Flow Scheduler (Dark themed screen)
Achieve lower implementation costs through recommended trade volumes that minimize market impact and price risk.
Collaborated with finance team to deliver trade flow scheduling product for one of the largest financial investment companies in the world. Provided product design strategy and visual design for this product.
Toolkit (AI Operations)
Orkestrator allows you to optimize the use of compute and storage resources, and bridge the gap between IT infrastructure and the needs of AI practitioners. Product design strategy was done for Toolkit which has an ensemble of tools that help AI practitioner run experiments, view the status of your jobs, build models and data label sets.
AI Generated Apparel Creator
An interactive application that allows a user to develop apparel concepts in concert with an AI model by exploring latent vectors (in multi-dimension latent space).
In collaboration with SSENSE ecommerce retailer, we were able to teach an AI model to generate almost infinite amounts of unique designs. When the project began the model would spit out thousands of images most were not good and a handful were nice but like a real creative process designs need to be iterated on (e.g., more blue, less patterns).
The theory was to design an application that allows the user/designer and AI to collaborate - allowing what the AI does best by generating almost infinite concepts in seconds and the user/designer to use their intuition on what looks nice, what styles work. The interactive application allows the designer to allow the AI and the designer have a back and forth workflow in order to get the best design concepts quickly.
We utilized the AI designs to produce bags and t-shirts and demonstrated our project and research at the NEURIPS (largest AI conference).
