mooflife-research | mooflife.com

<aside> 💡

Project Name - Wikipedia Revisions updater

Extend with - History-Aware Moment Creation Model

Project Type - Requirement gathering, development, deployment & maintenance

Duration - 3 months

Tech Stack

Python, Open AI, MongoDB, FastAPI, Qdrant Vector DB, Azure, GitHub Actions (CI/CD), Wikipedia

</aside>

Job Description

Developed and deployed application to capture the revisions happen in Wikipedia pages and parse those updates into a existing document database and update the vector databases in Qdrant Cloud.

Responsibilities and Accomplishments

Full Development Lifecycle Ownership:

• Managed the entire development lifecycle, from initial requirement gathering and system design to API development, deployment, and ongoing maintenance.
• Ensured seamless integration and functionality of the "Wikipedia Revision parser" system.

Advanced Revisions Parsing:

• Designed and implemented an advanced Revisions Parsing architecture, enabling the system to capture various types of revisions.

LLM Integration and Optimization:

• Utilized prompt engineering techniques to guide LLM outputs and enhance the quality of the push updates into existing “moments” with capturing the meaning.

Vector Database Management:

• Implemented and managed vector databases (Qdrant) for efficient storage and retrieval of historical data, update the existing vector database based on the revisions.
• Optimized vector search capabilities to improve the speed and accuracy of information retrieval.

Data Management and Integration: