app.py

import streamlit as st
import plotly.graph_objects as go
from resume_parser import extract_text
from resume_to_json import parse_resume
from ats import ats_score
from interview import generate_questions, evaluate_answer
import plotly.express as px


st.set_page_config(
    page_title="AI Interview Assistant",
    page_icon="🤖",
    layout="wide"
)

# ---------------- Sidebar ---------------- #

st.sidebar.title("🤖 AI Interview Assistant")

st.sidebar.markdown("""
### Features

- 📄 Resume Parsing
- 🎯 ATS Score
- 💼 Skill Match Analysis
- ❓ AI Interview Questions
- ⭐ AI Answer Evaluation

---
Built using:
- Streamlit
- Gemini AI
- PyMuPDF
""")

if st.sidebar.button("🔄 Start New Analysis"):
    st.session_state.clear()
    st.rerun()

hide_streamlit_style = """
<style>
#MainMenu {visibility: hidden;}
footer {visibility: hidden;}
header {visibility: hidden;}
</style>
"""

st.markdown(hide_streamlit_style, unsafe_allow_html=True)

# ---------------- Main UI ---------------- #

st.title(
    "🤖 Smart AI Interview Assistant"
)

st.caption(
    "Resume Analysis • ATS Matching • AI Mock Interview"
)

resume_file = st.file_uploader(
    "Upload Resume (PDF)",
    type=["pdf"]
)

job = st.text_area("Paste Job Description")

if not resume_file:

    st.info("📄 Please upload your resume.")

elif not job.strip():

    st.info("💼 Please paste a job description.")

else:

    st.success(f"Uploaded: {resume_file.name}")

    with open("temp_resume.pdf", "wb") as f:
        f.write(resume_file.read())

    # ---------- Resume Parsing ---------- #

    if "resume_json" not in st.session_state:

        with st.spinner("Analyzing resume..."):

            resume = extract_text("temp_resume.pdf")

            st.session_state.resume_json = parse_resume(resume)

    resume_json = st.session_state.resume_json

    st.subheader("📋 Resume Summary")

    c1, c2, c3 = st.columns(3)

    with c1:
        st.metric(
            "Skills",
            len(resume_json.get("skills", []))
        )
    
    with c2:
        st.metric(
            "Projects",
            len(resume_json.get("projects", []))
        )
    
    with c3:
        st.metric(
            "Education",
            len(resume_json.get("education", []))
        )
    
    with st.expander("View Parsed Resume JSON"):
        st.json(resume_json)

    # ---------- ATS ---------- #

    if "ats_result" not in st.session_state:

        with st.spinner("Calculating ATS Score..."):

            st.session_state.ats_result = ats_score(
                resume_json,
                job
            )

    result = st.session_state.ats_result

    st.subheader("🎯 ATS Score")

    fig = go.Figure(
        go.Indicator(
            mode="gauge+number",
            value=result["score"],
            number={"suffix": "%"},
            title={"text": "ATS Score"},
            gauge={
                "axis": {"range": [0, 100]},
                "bar": {"color": "green"}
            }
        )
    )

    st.plotly_chart(fig, use_container_width=True)

    score = result["score"]

    if score >= 80:
        st.success("🏆 Excellent Resume")
    elif score >= 60:
        st.warning("👍 Good Resume")
    else:
        st.error("⚠ Needs Improvement")

    col1, col2 = st.columns(2)

    with col1:
        st.success("Matched Skills")
        for skill in result["matched_skills"]:
            st.write(f"✅ {skill}")

    with col2:
        st.error("Missing Skills")
        for skill in result["missing_skills"]:
            st.write(f"❌ {skill}")

    report = f"""
ATS SCORE: {result['score']}%

Matched Skills:
{result['matched_skills']}

Missing Skills:
{result['missing_skills']}
"""

    fig = px.pie(
        names=["Matched", "Missing"],
        values=[
            len(result["matched_skills"]),
            len(result["missing_skills"])
        ],
        hole=0.6
    )

    st.plotly_chart(fig, use_container_width=True)

    st.download_button(
    "📥 Download ATS Report",
    report,
    file_name="ATS_Report.txt"
)

# ---------- Interview ---------- #

if st.button("Generate Interview Questions"):

    try:

        with st.spinner("Generating Interview Questions..."):

            st.session_state.questions = generate_questions(
                resume_json,
                job
            )["questions"]

    except Exception:

        st.error(
            "Unable to generate interview questions right now."
        )

# ---------- Interview ---------- #

if "questions" in st.session_state:

    st.header("🎤 AI Interview")

    for i, q in enumerate(st.session_state.questions):

        st.divider()

        st.subheader(
            f"🎯 Technical Question {i+1}"
        )
        st.info(q)

        ans = st.text_area(
            "Your Answer",
            key=f"ans{i}",
            height=150
        )

        if st.button(
            "Evaluate Answer",
            key=f"btn{i}"
        ):

            if not ans.strip():
                st.warning("Please type your answer before evaluation.")
                continue

            try:

                with st.spinner("Evaluating your answer..."):

                    evaluation = evaluate_answer(q, ans)

                st.metric(
                    label="Interview Score",
                    value=f"{evaluation['score']}/10"
                )

                with st.expander(
                    "📝 Feedback",
                    expanded=True
                ):
                    st.write(
                        evaluation["feedback"]
                    )

                with st.expander(
                    "💡 Ideal Answer"
                ):
                    st.write(
                        evaluation["ideal_answer"]
                    )

            except Exception:

                st.error(
                    "Unable to evaluate your answer right now. "
                    "Please try again later."
                )

# ---------- Footer ---------- #

st.divider()

st.caption(
    "Developed by Avani Rathi • AI Interview Assistant • Powered by Gemini AI"
)