Parth Shah

Experience

• 2024 - Present

  AI Engineering Intern @ Quantera.ai

  Stockholm, Sweden - Remote

  Summary:

  • Developed data visualization and automation solutions for financial analysis and reporting

  Responsibilities:

  • Built a visualization engine to extract data, figure plot requirements and generate code to build plots from relevant sources for question-answer pairs. Served this as a single FastAPI endpoint to frontend.
  • Developed cron jobs to automate daily ETL processes for updating Top 500 US companies' earnings calls.
  • Collaborated with cross-functional teams to enhance data-driven insights for financial analysis.

  • LlamaIndex
  • FastAPI
  • OLlama
  • AWS
• 2024 - Present

  Research Engineer @ Data Mining and Reinforcement Learning Lab

  Tempe, Arizona - On site

  Summary:

  • Working as a research intern building GraphRAG applications for PDF Question Answering

  Responsibilities:

  • Collaborated with Prof. Hua Wei and Longchao Da, contributing to the successful acceptance of our paper at the Agent4IR workshop, part of the 33rd ACM Conference on Information and Knowledge Management (CIKM 2024).
  • Developed EvidenceChat, a framework enhancing LLMs with Retrieval-Augmented Generation (RAG) and Chain-of-Thought (CoT) prompting to ensure trustworthy and evidential response generation.
  • Engaged in discussions and shared insights with fellow researchers and professionals to drive innovation in AI and information retrieval.

  • Python
  • Ollama
  • Streamlit
  • Git
• 2024 - 2024

  Research Engineering Intern @ Ethical Spectacle Research

  Tempe, Arizona - On site

  Summary:

  • Designed and deployed a backend for the SADIE platform, conducted simulations to analyze network effects, and developed algorithms for AI ethics research, enhancing data accuracy and reliability.

  Responsibilities:

  • Designed and deployed a robust backend for the SADIE platform, enabling seamless interaction and information storage across multiple language model agents on Weaviate.
  • Orchestrated simulations with hundreds of agents to analyze emergent network effects, improving data accuracy and reliability by 15% for research.
  • Developed algorithms to filter toxic and promote constructive ideas, contributing to AI ethics research.

  Achievements:

  • Python
  • Weaviate
  • DSPy
  • LangChain