Sigmoidal Success Story
7.1M+ reduction in overhead costs with AI-based Investment Signal Tracking Engine
357M+
Documents
and Posts Processed
7.1M+
Reduction
in overhead costs
10.9%+
Faster reaction
to market shifts
2020
Go-Live date
What was the business problem?
Baillie Gifford is an investment management firm headquartered in the UK, with over 1,200 employees, multiple offices worldwide (i.a. New York, Zurich, Hong Kong), and £262 billion (approx. $334 billion) in assets under management as of June 2020. More than 110 years on the market (founded in 1908). Baillie Gifford reached out to Sigmoidal, as they have been searching for proven solutions to adapt to the current digital transformation trends across the industry.
The project’s goal was to automate the process of sentiment tracking and analysis, enabling Baillie Gifford to make better-informed decisions concerning investment and asset allocation across their portfolio.
What savings and profits does the client achieve?
Sigmoidal augmented the process, significantly reducing the workload needed for research and analysis. We enhanced decision-making procedures as Baillie Gifford can now use additional investment signals to their advantage, getting it directly from users. Sigmoidals’ solution can automatically generate a high-level report, allowing managers to react to market behavior faster by an estimated 10.9%. It effectively reduced our client costs by a conservative number of 9.1M annually.
How did we accomplish it?
We designed and developed an AI-based tool to track and analyze clients’ sentiment towards companies from their portfolio, working in real-time. The cloud-based machine learning system core is a web application that provides aggregated metrics about clients’ posts from social media concerning their sentiment towards a company. It also automatically generates high-level reports, including only the crucial information. Among the others, we used spaCy, Docker, and TensorFlow as our tools. In addition, we used Twitter API for data collection and easy maintenance in the future (in case of any updates on this social media platform).
How did we boost the project with Sigmoidal DNA?
Design Authority
We utilized our proprietary Enterprise Machine Learning patterns to create the cloud-based machine learning system core, which provides aggregated metrics about clients’ posts from social media concerning their sentiment towards a company.
Product-focus Method
At the earliest phase, we quickly adapted our metrics definition and monitoring solutions to effectively put the project onto the production stage, allowing us to implement uniform data structures and begin versioning models early on – Minimizing the Total Cost of Ownership.
Knowledge Sharing
Sigmoidal team transferred the expertise and technological novelties to the client’s team, enabling easy maintenance and improvement in the future.
Deployment Strategy
Sigmoidal planned operation phases, understanding biases and sources of data. We monitored them to establish how the model’s accuracy could be improved with thoughtful feature engineering. We utilized this information to put the product onto the production cycle 1.8 months before the deadline.
What technical stack has been designed and implemented?
