Skip to content
View Sarvamm's full-sized avatar
7 followers · 7 following

Block or report Sarvamm

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Sarvamm/README.md

About Me

Data Science student specializing in AI and Machine Learning. Experienced in Python, SQL, Tableau. Built end-to-end AI apps for automation and data analysis.

Technical Skills

  • Languages: Python, C++
  • Libraries: Pandas, Numpy, Scikit-Learn, PyTorch, Langchain, Streamlit
  • Databases & Querying: SQL (PostgreSQL)
  • Data Visualization: Tableau, Power BI, Plotly, Seaborn, Matplotlib
  • Core Concepts: Data Structures & Algorithms, Databases, Exploratory Data Analysis, Machine Learning Pipelines

Experience

CSRBOX IBM Winter Internship — Prompt Engineer Intern
Dec. 2024 – Jan. 2025 · Remote
View Certificate

  • Designed an end-to-end product launch simulation by utilizing AI tools and advanced prompt engineering techniques.
  • Integrated artificial intelligence across key business functions, including product conceptualization, market analysis, marketing strategy, and financial planning.

Projects

Zeno 2 - Automated Data Analysis
Python, Streamlit, LangChain, Groq API, Jupyter Client, Plotly

  • Architected a stateful execution engine by embedding an isolated background Jupyter Kernel (jupyter_client), enabling safe runtime execution of AI-generated code while maintaining persistent variable state across chat turns.
  • Engineered a context-aware LLM agent using LangChain and Groq API (OSS 120B model), dynamically passing live kernel memory manifests and schema metadata to ensure precise code generation and minimal hallucination.
  • Integrated automated data profiling using a custom library (sinica) for immediate dataset summaries and anomaly alerts, alongside a dual-mode conversational UI supporting natural language queries and direct Python slash commands (/).
  • Built interactive visualization and export pipelines, streaming real-time execution logs, rendering interactive Plotly charts, and enabling full session compilation into native .ipynb Jupyter Notebooks.

Aurelius - Stoic AI Companion
Python, LangChain, ChromaDB, Streamlit, Groq

  • Developed an AI counselor channeling Marcus Aurelius, featuring dual-source retrieval across primary JSON aphorisms and historical PDF texts.
  • Implemented a Parent-Child (Small-to-Big) chunking strategy using BAAI/bge-m3 embeddings to perform vector searches on 500-character child passages while passing 2,500-character parent context to the LLM.
  • Designed a production-ready Streamlit interface with low-latency streaming inference, custom Cormorant Garamond typography, and a context inspector for transparent source attribution.

Certifications

Pinned Loading

  1. ds-ml-projects ds-ml-projects Public

    Mini projects intended to learn ML

    Jupyter Notebook 1

  2. Aurelius Aurelius Public

    Stoic RAG chatbot channeling Marcus Aurelius. Featuring a dual-source retrieval engine that queries primary JSON aphorisms alongside historical PDF texts via BGE-M3 embeddings. Built with parent-ch…

    Python

  3. Sinica Sinica Public

    Python package for quick data profiling and data quality alerts, inspired by y-data-profiling

    Python

  4. Zeno-2 Zeno-2 Public

    Zeno 2 is a data analysis platform powered by LLMs. Talk to your data in natural language, get interactive visualizations, run commands directly and export notebook when you're done.

    Python