In [1]:

def whoami():
    return {
        "name": "Sujay Pandey",
        "role": "Aspiring AI/ML Engineer",
        "status": "CSIT student, final semester",
        "education": "BSc. CSIT, Lumbini City College "
                     "(affiliated with Tribhuvan University), 2022–2026",
    }

whoami()

Out[1]:

Sujay Pandey

Aspiring AI/ML Engineer · CSIT Student

Open to internship opportunities

education: BSc. CSIT, Lumbini City College (affiliated with Tribhuvan University), 2022–2026

download resume

In [2]:

""" About me — rendered below. """

Out[2]:

I'm in my final (8th) semester of BSc. CSIT, where the coursework built a foundation in C, C++, Java, JavaScript, .NET, SQL and databases. Outside of that, I've been teaching myself Python and ML/AI.

Focused right now on building small end-to-end ML projects — data handling, model building, and evaluation — starting with a movie recommendation system.

projects/

In [3]:

project_1 = Project(
    name="RAG-Resume-Matcher",
    type="RAG-based resume-to-JD matching tool",
    stack=["Django", "Sentence-Transformers", "ChromaDB", "Gemini API"],
    status="completed"
)
project_1.display()

Out[3]:

RAG-Resume-Matcher — evidence-backed resume-to-job-description analysis.

Built a tool that matches resumes against job descriptions and generates an evidence-backed narrative explaining strengths and gaps. Retrieval pipeline: PDF/DOCX parsing, section-aware chunking, sentence-transformer embeddings, and ChromaDB vector storage. Two-stage multi-pass LLM pipeline (requirement extraction → evidence mapping → self-check → narrative summary) with Gemini-primary/Groq-fallback, served via Django REST API. Includes multi-user support with per-session vector-store isolation and a DB-backed daily usage cap.

→ view repo → live demo

In [4]:

project_2 = Project(
    name="Movie-Recommendation-Engine",
    type="content-based recommendation system",
    stack=["pandas", "scikit-learn"],
    status="completed"
)
project_2.display()

Out[4]:

Movie-Recommendation-Engine — content-based movie recommendation system.

Movie-Recommendation-Engine — Full-stack movie recommender (FastAPI + React) using precomputed sentence-transformer embeddings (all-MiniLM-L6-v2), vectorized cosine similarity, fractional genre-overlap scoring, and a logarithmic popularity weight — across a 49k-title TMDB dataset. Includes typo-tolerant search and a 30-test suite.

→ view repo → live demo

In [5]:

project_3 = Project(
    name="Sirupata-farm-house",
    type="Django-based hotel web application",
    stack=["Django", "MySQL", "scikit-learn", "TextBlob"],
    status="completed"
)
project_3.display()

Out[5]:

Sirupata-farm-house — Django-based hotel website with ML recommendations.

Sirupata Farm House — Full-stack Django hotel booking platform for a farm-led retreat, with a dedicated ml_engine app handling content-based room recommendations (TF-IDF + cosine similarity, scikit-learn) and review sentiment scoring (TextBlob polarity blended with star ratings). Includes a booking-operations dashboard (monthly trends, room popularity, check-in/out tracking), AJAX-driven review sorting, and Docker/MySQL deployment.

→ view repo → live demo

In [6]:

$ pip list --group

Out[6]:

Machine Learning

python
pandas
scikit-learn

Core Languages

c / c++
java

Web / Tools

javascript
html / css
sql
.net

# currently installing: transformers, deep learning fundamentals

In [7]:

# cat certifications.json

Out[7]:

CS50P — Introduction to Programming with PythonHarvard (CS50)2026

→ verify

PandasKaggleMay 21, 2026

→ verify

In [8]:

def contact_me():
    return ["email", "github", "linkedin"]

contact_me()

Out[8]:

→ email me → GitHub → LinkedIn