endorsedCompleted

Financial Document Analysis Engine (AI Training Libraries)

Services Covered:

Custom Software Development • AI and Machine Learning

Industry

Financial Services

Duration

3 months

Budget

10K - 49K

Start Date

1 January 2020

Team Size

1-5 Employees

Engagement Model

Offshore

Project Capability Score (PCS)

73/100

Good

Project Capability Score (PCS) estimates how capable WalkingTree Technologies is of successfully delivering a project like this, based on its past experience, track record, and ability to handle similar work. Learn more about PCS

Client

Confidential

New York City, United States of America

Project Summary

WalkingTree collaborated with an investment firm's technical lead to develop software libraries and scripts to train AI on financial documents, including parsing data, building training datasets, running named entity recognition and creating a Q&A mechanism. A three-person team (project manager, developer and founder) delivered the initial beta-focused components between January and April 2020. The client extended the delivered library for internal use and praised the team's communication and technical competence.

Key Challenges

Develop libraries to parse downloaded financial data
Create a training dataset from parsed data
Run named entity recognition on the data
Build a Q&A mechanism for financial documents and reach a beta product

Project Deliverables

Data-parsing libraries
Training-dataset creation scripts
Named entity recognition components
Q&A mechanism for financial documents

Project Solution

WalkingTree provided initial software development and library implementation to parse downloaded financial data, create training datasets, run named entity recognition (NER) and develop a Q&A mechanism for financial documents. The small team (one project manager, one developer and the founder) used Python and NLP tools (Spacy, Huggingface) and managed communication via Slack, email and calls.

Project Outcome

The client extended the library developed for their own internal use
The client reported the work was high quality and the team was easy to approach and competent
Effective communication via Slack, email and calls facilitated progress

Platforms

AI/ML Platform
Web

Tech Stack

Python

Client Endorsement

Overall Review Rating

Timeliness

Cost Rating

Willing to Refer

Quality of Deliverables

“The company was so satisfied with the team's work that the company extended the library that they developed for their own use. The company was most impressed by the high-level communication that the team used during the project that allowed them to move forward.”
Anonymous
Investment Banker
Date: 09/16/2020