Update app.py

app.py

@@ -1,138 +1,431 @@
 import gradio as gr
+import google.generativeai as genai
+import os
 import random
-from google import generativeai as genai
+import whisper
 
-# Configure Gemini
-genai.configure(api_key="YOUR_API_KEY_HERE")
-model = genai.GenerativeModel("gemini-2.0-flash")
+# ========== Gemini API Setup (fallback with 2 keys) ==========
+API_KEYS = [os.getenv("API_KEY_1"), os.getenv("API_KEY_2")]
+model = None
 
-# Evaluation logic
-def getscore(question, answer):
-    info = f"""
-You are an interview expert but be a little free not toooo strict. A user was asked the following interview question:
+def initialize_model():
+    global model
+    for key in API_KEYS:
+        try:
+            genai.configure(api_key=key)
+            model = genai.GenerativeModel("gemini-2.0-flash")
+            test = model.generate_content("Hello!")
+            if "Hello" in test.text:
+                print(f"[INFO] Using Gemini API key ending with: ...{key[-4:]}")
+                return
+        except Exception as e:
+            print(f"[WARNING] Gemini key failed: {e}")
+    raise RuntimeError("All Gemini API keys failed.")
 
-Question: {question}
-
-Their answer was: "{answer}"
-
-Please do the following:
-1. Rate the answer from 1 to 10 based on relevance, clarity, and depth.
-2. Provide detailed feedback on how the answer can be improved in a single line.
-Respond in the format:
-1.  **Rating:** <score>/10
-2.  **Feedback:** <your feedback here>
-"""
-    response = model.generate_content(info)
-    text = response.text.strip()
-    lines = text.split("\n")
-    score = -1
-    feedback = "Feedback not found"
-    for line in lines:
-        if "**Rating:**" in line:
-            try:
-                score = int(line.split("**Rating:**")[1].split("/")[0].strip())
-            except:
-                score = -1
-        elif "**Feedback:**" in line:
-            feedback = line.split("**Feedback:**")[1].strip()
-    return [score, feedback]
-
-def finalsummary(feedback_list):
-    prompt = f"""
-You are given a list of feedback. Summarize it and give 5–6 final suggestions for improvement.
-feedback={feedback_list}
-"""
-    response = model.generate_content(prompt)
-    return response.text
+initialize_model()
 
-# Question bank (shortened example)
-imp_questions_map = {
+# ========== Questions ==========
+all_questions = {
+    "general": [
+    "Tell me about yourself.",
+    "What are your strengths and weaknesses?",
+    "How do you handle stress or pressure?",
+    "Describe a time when you worked in a team.",
+    "How do you manage your time and prioritize tasks?",
+    "Why do you want to join this company?",
+    "Where do you see yourself in 5 years?",
+    "Describe a challenge or conflict you faced and how you dealt with it.",
+    "What motivates you to do your best?",
+    "How do you handle feedback and criticism?",
+    "Tell me about a time you took initiative.",
+    "How do you stay organized during busy schedules?",
+    "What do you do when you face a task you don’t know how to solve?",
+    "Tell me about a mistake you made and what you learned from it.",
+    "Have you ever led a project or team? What was your experience?",
+    "How do you manage deadlines when working on multiple tasks?",
+    "Why should we hire you?",
+    "Tell me about a time you failed. How did you respond?",
+    "How do you approach learning something new?",
+    "What do you know about our company?",
+    "How do you balance quality and speed in your work?",
+    "Describe a time when you had to make a difficult decision.",
+    "How do you communicate in a remote or hybrid work environment?",
+    "What kind of work culture are you most comfortable in?",
+    "Do you prefer working independently or in a team? Why?"
+    ],
+    "web development": [
+        "What is the difference between HTML and XHTML?",
+        "Explain how RESTful APIs work.",
+        "What are the advantages of using React over plain JavaScript?",
+        "How does the virtual DOM improve performance?",
+        "What are HTTP methods? Explain each.",
+        "What is the difference between sessionStorage and localStorage?",
+        "What is CORS and how do you handle it?",
+        "Explain the box model in CSS.",
+        "What is responsive design?",
+        "What is the difference between SQL and NoSQL databases?",
+        "What are web sockets and when are they used?",
+        "What is a service worker in PWA?",
+        "What is lazy loading and why is it useful?",
+        "Difference between GET and POST requests?",
+        "How does server-side rendering differ from client-side rendering?",
+        "What is Next.js and how is it different from React?",
+        "What is a CDN and how does it help?",
+        "What is HTTPS and why is it important?",
+        "Explain the concept of middleware in Express.js.",
+        "What is an ORM and why use it?",
+        "Explain MVC architecture.",
+        "What is the difference between PUT and PATCH requests?",
+        "What is GraphQL and how does it compare to REST?",
+        "How do cookies and sessions work?",
+        "What is JWT and how is it used in authentication?"
+    ],
+    "devops": [
+        "What is CI/CD and why is it important?",
+        "Explain the concept of Infrastructure as Code.",
+        "What is Docker and how is it different from a virtual machine?",
+        "What are containers and how do they help in deployment?",
+        "What is Kubernetes used for?",
+        "How do you monitor applications in production?",
+        "What is the role of Jenkins in DevOps?",
+        "What are some popular logging tools?",
+        "Explain blue-green deployment.",
+        "What is Ansible used for?",
+        "What is Terraform and how is it used?",
+        "What is a build pipeline?",
+        "What is Helm in Kubernetes?",
+        "What is a rolling update in deployments?",
+        "Explain canary deployments.",
+        "What is load balancing and how does it work?",
+        "What is the difference between horizontal and vertical scaling?",
+        "What are artifacts in a CI/CD pipeline?",
+        "What is the use of Prometheus and Grafana?",
+        "What is a reverse proxy?",
+        "How do you ensure zero-downtime deployments?",
+        "Explain the role of GitOps.",
+        "What are secrets management tools?",
+        "What is the purpose of using NGINX in DevOps?",
+        "What is a container registry?"
+    ],
+    "cybersecurity": [
+        "What is the difference between symmetric and asymmetric encryption?",
+        "What is SQL injection and how do you prevent it?",
+        "What are some common types of malware?",
+        "What is the CIA triad?",
+        "What is multi-factor authentication?",
+        "What are firewalls and how do they work?",
+        "Explain HTTPS and how it secures data.",
+        "What is penetration testing?",
+        "How do you secure an API?",
+        "What are some best practices for password security?",
+        "What is phishing and how can it be prevented?",
+        "What is social engineering?",
+        "What is a DDoS attack?",
+        "What are intrusion detection systems?",
+        "What is a VPN and how does it work?",
+        "What is data encryption at rest vs in transit?",
+        "What is a zero-day vulnerability?",
+        "Explain role-based access control (RBAC).",
+        "What is a digital certificate?",
+        "How does a public key infrastructure (PKI) work?",
+        "What is a man-in-the-middle attack?",
+        "What are honeypots in cybersecurity?",
+        "Explain sandboxing in malware analysis.",
+        "What are security patches?",
+        "What is OWASP Top 10?"
+    ],
+    "ai": [
+        "What is the difference between AI, ML, and Deep Learning?",
+        "Explain a neural network and how it learns.",
+        "What is gradient descent?",
+        "What is the role of activation functions in neural networks?",
+        "Explain supervised vs. unsupervised learning.",
+        "What is backpropagation?",
+        "What is overfitting and how do you avoid it?",
+        "What is transfer learning?",
+        "What are some common use-cases of AI?",
+        "How do you evaluate a classification model?",
+        "What is reinforcement learning?",
+        "What is an autoencoder?",
+        "Explain convolutional neural networks (CNNs).",
+        "What are recurrent neural networks (RNNs)?",
+        "What is a transformer model?",
+        "What is attention mechanism in NLP?",
+        "What is the Turing test?",
+        "What are generative adversarial networks (GANs)?",
+        "What is the bias-variance tradeoff?",
+        "What is explainable AI (XAI)?",
+        "How do you train a deep learning model?",
+        "What is early stopping in training?",
+        "What are hyperparameters?",
+        "What is a confusion matrix?",
+        "What is the role of regularization in deep learning?"
+    ],
     "ml": [
         "What is supervised learning? Give an example.",
         "What is the difference between classification and regression?",
         "How does linear regression work?",
         "What are decision trees and how do they work?",
-        "What is overfitting vs underfitting?"
+        "What is overfitting vs underfitting?",
+        "Explain cross-validation and its importance.",
+        "What is regularization in machine learning?",
+        "What is the difference between bagging and boosting?",
+        "What is the role of a cost function?",
+        "How do you handle missing data in ML datasets?",
+        "What is the k-nearest neighbors (KNN) algorithm?",
+        "What is support vector machine (SVM)?",
+        "What is Naive Bayes algorithm?",
+        "What is logistic regression?",
+        "What is feature scaling and why is it needed?",
+        "What is principal component analysis (PCA)?",
+        "What is model drift?",
+        "How do you tune hyperparameters?",
+        "What is grid search and random search?",
+        "What are ensemble models?",
+        "What is a confusion matrix and what do its values represent?",
+        "How do you balance an imbalanced dataset?",
+        "What is precision-recall tradeoff?",
+        "What is ROC-AUC score?",
+        "What is the purpose of stratified sampling?"
     ],
-    "general": [
-        "Tell me about yourself.",
-        "What are your strengths and weaknesses?",
-        "How do you handle stress or pressure?",
-        "Describe a time when you worked in a team.",
-        "Why should we hire you?"
+    "mobile development": [
+        "What is the difference between native and hybrid apps?",
+        "Explain the lifecycle of an Android activity.",
+        "What is Flutter and why use it?",
+        "How do you handle data persistence in mobile apps?",
+        "What is MVVM architecture?",
+        "Explain push notifications and how they work.",
+        "What are the limitations of mobile development?",
+        "How do you optimize performance in mobile apps?",
+        "What is the role of App Store Optimization (ASO)?",
+        "Explain the concept of responsive design in mobile.",
+        "What is the difference between Android and iOS development?",
+        "What is React Native and how does it work?",
+        "How do you handle offline mode in apps?",
+        "What is a RecyclerView?",
+        "Explain intents in Android.",
+        "What is dependency injection in Android?",
+        "What are permissions in mobile apps?",
+        "How do you debug a mobile app?",
+        "What is Jetpack Compose?",
+        "How is state management handled in Flutter?",
+        "What is a storyboard in iOS?",
+        "What are background services?",
+        "How do you handle API calls in mobile apps?",
+        "What is the difference between LiveData and Flow?",
+        "How do you publish an app to Google Play Store?"
+    ],
+    "uiux": [
+        "What is the difference between UI and UX?",
+        "What are some principles of good UX design?",
+        "What is a user journey?",
+        "How do you conduct user testing?",
+        "What tools do you use for UI/UX design?",
+        "What is wireframing?",
+        "How do you design for accessibility?",
+        "What is the role of prototyping?",
+        "Explain the concept of design thinking.",
+        "How do you collect and incorporate user feedback?",
+        "What is the importance of consistency in UI design?",
+        "What is heuristic evaluation?",
+        "What is the role of typography in UI?",
+        "What are color theory principles?",
+        "How does UX impact business goals?",
+        "What is A/B testing in UI/UX?",
+        "What is microcopy and why is it important?",
+        "Explain responsive vs adaptive design.",
+        "What is a user persona?",
+        "What is interaction design?",
+        "What is a high-fidelity vs low-fidelity prototype?",
+        "How do you measure UX success?",
+        "What is card sorting?",
+        "What are accessibility standards (WCAG)?",
+        "What is the role of animation in UI?"
+    ],
+    "data analytics": [
+        "What is the difference between descriptive and inferential statistics?",
+        "What are some common data cleaning techniques?",
+        "Explain how to deal with missing data.",
+        "What is the role of data visualization?",
+        "What is a pivot table?",
+        "What are outliers and how do you handle them?",
+        "Explain correlation vs causation.",
+        "What are some commonly used data analytics tools?",
+        "How do you handle large datasets efficiently?",
+        "What is ETL and how does it work?",
+        "What is a histogram and when is it used?",
+        "What is data wrangling?",
+        "What is the difference between mean, median, and mode?",
+        "What is standard deviation and variance?",
+        "What is a box plot?",
+        "What is a time series analysis?",
+        "What is the difference between R and Python for analytics?",
+        "What is the use of SQL in data analytics?",
+        "How do dashboards help in analytics?",
+        "What is cohort analysis?",
+        "What is hypothesis testing?",
+        "What are confidence intervals?",
+        "How do you avoid bias in analytics?",
+        "What is the purpose of a heatmap?",
+        "What is a funnel analysis?"
+    ],
+    "data scientist": [
+        "What is the difference between supervised and unsupervised learning?",
+        "Explain bias-variance tradeoff in machine learning.",
+        "How do you handle imbalanced datasets?",
+        "What are precision, recall, and F1-score? Why are they important?",
+        "Describe a machine learning project you've worked on.",
+        "What is feature engineering? Give examples.",
+        "How do you select important variables in a dataset?",
+        "Explain overfitting and how to prevent it.",
+        "What’s the difference between bagging and boosting?",
+        "Describe the steps in a typical machine learning pipeline.",
+        "What tools do you use as a data scientist?",
+        "What is dimensionality reduction?",
+        "How do you evaluate regression models?",
+        "What is the difference between ROC and PR curves?",
+        "What is model interpretability?",
+        "How do you deploy a machine learning model?",
+        "What is a recommendation system?",
+        "How do you deal with outliers?",
+        "What is SMOTE?",
+        "What is A/B testing in data science?",
+        "How do you build a data pipeline?",
+        "What are data lakes and data warehouses?",
+        "What is a baseline model?",
+        "How do you perform EDA (Exploratory Data Analysis)?",
+        "What is clustering and what are its applications?"
+    ],
+    "prompt engineering": [
+        "What is prompt engineering and why is it important?",
+        "How do temperature and top_p affect model output in LLMs?",
+        "How do you write a good system prompt for a chatbot?",
+        "What are the risks of prompt injection attacks?",
+        "What is chain-of-thought prompting?",
+        "How does few-shot prompting improve performance?",
+        "What are embeddings, and how are they used in retrieval-augmented generation?",
+        "What are the limitations of prompt-based techniques?",
+        "Explain the role of role-based prompts in function calling.",
+        "How would you structure a prompt to summarize a large document accurately?",
+        "What is zero-shot learning in prompt design?",
+        "What is in-context learning?",
+        "What is semantic search and how is it used with LLMs?",
+        "How do you evaluate the output of an LLM?",
+        "What is hallucination in LLMs?",
+        "What are guardrails in LLM applications?",
+        "What is prompt chaining?",
+        "How do you personalize prompts for users?",
+        "What are system vs user prompts?",
+        "What is instruction tuning?",
+        "How do you manage memory in prompt-based agents?",
+        "What is RAG (retrieval-augmented generation)?",
+        "What is the difference between deterministic and probabilistic LLM output?",
+        "How do prompts interact with tools in agent systems?",
+        "How do you test prompt robustness?"
     ]
 }
 
-# App state
+# ========== Whisper ASR ==========
+whisper_model = whisper.load_model("base")
+
+# ========== Session State ==========
 session = {
     "questions": [],
     "current_index": 0,
-    "answers": [],
     "scores": [],
-    "feedbacks": []
+    "feedbacks": [],
+    "transcripts": []
 }
 
-# Main logic handler
-def start_session(field):
-    session["questions"] = random.sample(imp_questions_map.get(field, []), 5) + random.sample(imp_questions_map["general"], 3)
-    session["current_index"] = 0
-    session["answers"] = []
-    session["scores"] = []
-    session["feedbacks"] = []
-    return session["questions"][0]
+# ========== Gemini Evaluation ==========
+def get_score(question, answer):
+    prompt = f"""
+You are an interview expert. A user was asked:
+
+Question: {question}
+
+Their answer was: "{answer}"
 
-def submit_answer(audio):
-    import whisper
-    model = whisper.load_model("base")
-    audio_path = audio
-    result = model.transcribe(audio_path)
-    answer = result["text"]
+Please:
+1. Rate from 1 to 10 based on clarity, relevance, and depth.
+2. Give a 1-line constructive feedback.
+
+Format:
+1. **Rating:** <score>/10
+2. **Feedback:** <text>
+"""
+    response = model.generate_content(prompt).text.strip()
+    score = -1
+    feedback = "Feedback not found"
+    for line in response.split("\n"):
+        if "**Rating:**" in line:
+            try:
+                score = int(line.split("**Rating:**")[1].split("/")[0].strip())
+            except:
+                pass
+        elif "**Feedback:**" in line:
+            feedback = line.split("**Feedback:**")[1].strip()
+    return score, feedback
+
+# ========== Summary ==========
+def summarize_feedback(feedbacks):
+    prompt = "Summarize the following feedback into 4-5 improvement tips:\n" + "\n".join(feedbacks)
+    return model.generate_content(prompt).text.strip()
+
+# ========== Audio Transcription ==========
+def handle_audio(audio):
+    if not audio:
+        return "No audio provided"
+    result = whisper_model.transcribe(audio)
+    return result["text"]
+
+# ========== Process Each Answer ==========
+def run_session(audio):
+    if session["current_index"] >= len(session["questions"]):
+        return "✅ Interview complete!", gr.update(visible=False)
 
-    # Evaluate
     question = session["questions"][session["current_index"]]
-    score, fb = getscore(question, answer)
-    session["answers"].append(answer)
-    session["scores"].append(score)
-    session["feedbacks"].append(fb)
+    transcript = handle_audio(audio)
+    session["transcripts"].append(transcript)
 
+    score, feedback = get_score(question, transcript)
+    session["scores"].append(score)
+    session["feedbacks"].append(f"Q{session['current_index']+1}: {feedback}")
     session["current_index"] += 1
 
     if session["current_index"] < len(session["questions"]):
         next_q = session["questions"][session["current_index"]]
-        return next_q, gr.update(visible=True), gr.update(visible=False), ""
+        return f"🗣️ Answer: {transcript}\n\n🔢 Score: {score}/10\n💬 Feedback: {feedback}\n\n➡️ Next: {next_q}", gr.update(visible=True)
     else:
-        # Done - show summary
-        total_score = sum(session["scores"]) / len(session["scores"])
-        summary = finalsummary(session["feedbacks"])
-        return "Interview completed!", gr.update(visible=False), gr.update(visible=True), f"Average Score: {total_score:.2f}/10\n\nSummary:\n{summary}"
-
-def show_detailed_feedback():
-    detailed = ""
-    for i, q in enumerate(session["questions"]):
-        detailed += f"**Q{i+1}: {q}**\nAnswer: {session['answers'][i]}\nScore: {session['scores'][i]}/10\nFeedback: {session['feedbacks'][i]}\n\n"
-    return detailed
-
-# Gradio UI
-with gr.Blocks() as demo:
-    gr.Markdown("## AI Mock Interview Evaluator")
-
-    field_input = gr.Dropdown(
-        choices=list(imp_questions_map.keys()),
-        label="Select your field"
-    )
-
-    start_btn = gr.Button("Start Interview")
-    question_output = gr.Textbox(label="Question", interactive=False)
-    audio_input = gr.Audio(source="microphone", type="filepath", label="Record your answer")
-    submit_btn = gr.Button("Submit Answer")
-
-    final_output = gr.Textbox(label="Final Score and Summary", visible=False)
-    detailed_btn = gr.Button("Show Detailed Feedback", visible=False)
-    detailed_output = gr.Markdown()
-
-    start_btn.click(start_session, inputs=field_input, outputs=question_output)
-    submit_btn.click(submit_answer, inputs=audio_input,
-                     outputs=[question_output, audio_input, final_output, detailed_output])
-    detailed_btn.click(show_detailed_feedback, outputs=detailed_output)
+        avg_score = sum(session["scores"]) / len(session["scores"])
+        summary = summarize_feedback(session["feedbacks"])
+        return f"✅ Interview Complete!\n\n📊 Avg Score: {avg_score:.2f}/10\n\n📝 Summary:\n{summary}", gr.update(visible=False)
+
+# ========== Start New Session ==========
+def start_session(field):
+    general_qs = random.sample(all_questions["general"], 10)
+    domain_qs = random.sample(all_questions[field], 10)
+    session["questions"] = general_qs + domain_qs
+    random.shuffle(session["questions"])
+    session["current_index"] = 0
+    session["scores"] = []
+    session["feedbacks"] = []
+    session["transcripts"] = []
+    return session["questions"][0], gr.update(visible=True)
+
+# ========== Gradio UI ==========
+with gr.Blocks() as demo:web development
+    gr.Markdown("## 🎤 AI Mock Interview (Voice-Based) — 20 Questions")
+    with gr.Row():
+        field = gr.Dropdown(label="Select your domain", choices=["ml", "data science","web development","devops","cybersecurity","ai","mobile development","uiux","data analytics","prompt engineering"], value="data science")
+        start_btn = gr.Button("Start Interview")
+    
+    question_display = gr.Textbox(label="Current Question", interactive=False)
+    audio_input = gr.Audio(type="filepath", label="🎙️ Speak your answer")
+    submit_btn = gr.Button("Submit Answer", visible=False)
+    output = gr.Textbox(label="Feedback", lines=8)
+
+    start_btn.click(start_session, inputs=field, outputs=[question_display, submit_btn])
+    submit_btn.click(run_session, inputs=audio_input, outputs=[output, submit_btn])
 
 demo.launch()