Disqover Assistant (Artificial Intelligence System)

Python AWS & AI Expert at a tech consulting company

100 people affected

10 people managed

9 month project

• As a Python & AI expert I contributed to build and AI Assistant build on the top

of Disqover product which is used for research in Health Care Domain for

example clinical trials data, medical publications etc. this product is used by

clients like Johnson & Johnson, AstraZeneca

• DQ Assistant is based on Large Language Models which helps user to search

and explore during their research

• It provides capabilities like generation of DQ Query from natural language input

and summarizes unstructured text given by user.

Team Size : 5

Scripting Language : Python

Framework : OpenAI, ChatGPT, AWS, Microsoft Azure OpenAI, Flask

Tools Used : Kubernetes, Docker, Git, IntelliJ IDE, Dude, OpenLens

AI Techniques : LLM, Benchmarking, Embedding, Vectorization,

Prompt Engineering, Data Wrangling, Data Pip

Under budget

Steep learning curve