Skip to content
View sharpdima's full-sized avatar

Block or report sharpdima

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.

Maximum 250 characters. Please donโ€™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 userโ€™s behavior. Learn more about reporting abuse.

Report abuse
sharpdima/README.md

Hi there, I'm Maryam Asgari ๐Ÿ‘‹

AI System Designer & Data Infrastructure Engineer @ Sharp Dima

I am a technical consultant and system designer specializing in scalable data models, infrastructure architecture, and applied AI. Currently, my engineering focus is on architecting highly accurate, multi-lingual Retrieval-Augmented Generation (RAG) systems and tailoring open-source platforms to meet complex enterprise and organizational needs.

๐Ÿš€ What I Do & Recent Highlights

  • Enterprise Intelligent Assistant Development: Leading the technical infrastructure for an advanced, proprietary AI assistant utilizing the DeepSeek Chat model, seamlessly integrating specialized research documents and complex API configurations.
  • Advanced RAG Architecture: Designing strict-prompt, zero-hallucination RAG pipelines for prominent research institutes and charity organizations. These systems are capable of processing and responding dynamically in Persian, Arabic, and English.
  • Intelligent Data Preprocessing: Developed a custom Python/Streamlit software for NLP document processing. The engine dynamically parses large datasets, applies hierarchical chunking (preserving context up to H5 levels), injects metadata, and generates targeted keywords per chunk to optimize vector search limits.
  • AI Platform Customization (Dify):
    • Backend Engineering: Rewrote chat-title generation logic for accurate Persian/Arabic context extraction.
    • Dynamic UI/UX & RTL Support: Engineered fully responsive LTR/RTL context switching (including Vazirmatn typography injection and layout mirroring) based on chat language.
    • Custom Verification Modules: Built an interactive layer allowing users to cross-reference AI answers with live internet searches, generating academic-style citations (e.g., [1], [2]) directly within the UI.
  • Scalable Data Systems: Designing and deploying robust server-integrated data models using Docker and Directus, featuring advanced validation rules and activity assessment modules.

๐Ÿ› ๏ธ Tech Stack & Tools

  • Languages: Python, JavaScript/TypeScript
  • AI & NLP: RAG Architecture, Prompt Engineering, LLM Optimization (DeepSeek)
  • Data & Infrastructure: Docker, Directus, Vector Databases, Environment Management
  • Frontend/Tools: Streamlit, Tailwind CSS, PyInstaller, API Integrations

๐Ÿ“ซ Let's Connect: sharpdigimarketing@gmail.com

Popular repositories Loading

  1. dify dify Public

    Forked from langgenius/dify

    Production-ready platform for agentic workflow development.

    TypeScript

  2. moodle moodle Public

    Forked from moodle/moodle

    Moodle - the world's open source learning platform

    PHP

  3. plane plane Public

    Forked from makeplane/plane

    ๐Ÿ”ฅ ๐Ÿ”ฅ ๐Ÿ”ฅ Open Source JIRA, Linear, Monday, and Asana Alternative. Plane helps you track your issues, epics, and cycles the easiest way on the planet.

    TypeScript

  4. noor_theme noor_theme Public

    PHP

  5. sharpdima sharpdima Public