# Project export: TalkThru This document was generated by HackStack to give an AI agent context about a hackathon project. Sections are labeled with their provenance; content marked as truncated was cut to keep this document small. ## Project metadata - Hackathon: TreeHacks 2025 - Tagline: Practice interviews & difficult conversations. - Devpost: https://devpost.com/software/talkthru - GitHub: https://github.com/Andrea-MiramonSerr/tree-hacks.git - Video: https://www.youtube.com/embed/9GqcxNZc7lU?enablejsapi=1&hl=en_US&rel=0&start=&version=3&wmode=transparent - Team: 2 GitHub contributor(s) — Nelson Ooi (7 commits), am2389@cornell.edu (6 commits) ## Devpost submission (written by the team) ### Overview Technical interviews, presentations, salary negotiations, sharing bad news, thesis defenses … what do all these have in common? You need to practice. You need feedback. TalkThru is the professional you need for on-demand, face-to-face practice. Technical or behavioral, emotional or cold: test your speeches, knowledge & improvisation with immediate corrections. You provide a topic to the agent. You provide the agent a role. The agent holds a conversation with you and summarizes what you should improve on at the end. It’s that easy. 5 steps: Program initialization: User provides role & theme for the interview. Query: a) Share the prompt with Perplexity to generate an introduction, b) generate a fitting video with LumaLabs. ElevenLabs: translate Perplexity's text-based response to speech & begin the conversation. Speech2Text: The user's response to the agent's question is translated to text (using python speech_recognition). Re-query: Unless the user wants to end the conversation, repeat steps 2-4. Speech-to-text: pause identification, fast responses, storing & processing audio real-time were challenging aspects of the speech integration. Speech-to-text: pause identification, fast responses, storing & processing audio real-time were challenging aspects of the speech integration. Lip-syncing: realistic interview preparation requires mouth movement. Whereas LumaLabs provides vastly creative features, it does not support facial expressions. Our first feature release will enable lip syncing so that users can carry realistic conversations. Lip-syncing: realistic interview preparation requires mouth movement. Whereas LumaLabs provides vastly creative features, it does not support facial expressions. Our first feature release will enable lip syncing so that users can carry realistic conversations. CUDA integration: lip-syncing requires torch access, CUDA-based GPUs & many pre-trained models. Thanks to our sponsor NVIDIA, we were able to develop this feature. CUDA integration: lip-syncing requires torch access, CUDA-based GPUs & many pre-trained models. Thanks to our sponsor NVIDIA, we were able to develop this feature. Markdown & React.js fans try flask for the first time. GenAI is actually pretty fast. Run LLMs on CUDA! Run LLMs on CUDA! End-to-end integration of LLM queries for on-demand user requests. End-to-end integration of LLM queries for on-demand user requests. Successful project scoping Successful project scoping There is so much ground to cover: [Sun 2/16] Simultaneous conversations [Sun 2/23] Interview Summary & Grading [Sun 2/23] Interview Summary & Grading [Sun 2/23] Programming interviews [Sun 3/2] Design questions uploading Sketches & Block diagrams [Sun 3/9] User Assesment - free product campaign [Sun 3/16] Enhance sentimentality [ Difficult Conversations] ## README (from the GitHub repository) No README available. ## Detected evidence (automated analysis) Indexed codebase: 12 recognized source files, 34 KB. - HTML (language) — detected in the code - JavaScript (language) — detected in the code - Python (language) — detected in the code - Flask (technology) — claimed on Devpost, not found in the code ## Codebase structure (from repository index) ### Files (17 of 17) ``` .DS_Store .vscode/launch.json app.py eltest.py lumatest.py main.py perplexitytest.py speechrecog.py static/.DS_Store static/audio/.DS_Store static/js/main.js templates/.DS_Store templates/about.html templates/index.html templates/result.html tester.py texttosspeech.py ``` ### Dependencies No dependency index available. ### Recent commits (newest first) - Working version - Ported over functionality to flask app - Record conversations in webapp - prompt engineering - Can have conversation with thing - Merge pull request #1 from Andrea-MiramonSerr/feature/perplexity-tester - Confirmed follow-up responses & implemented memory. Confirming memory needs for continuous calling of the ask_perplexity function. This will require limiting history/summarizing. Ready to merge to main. - Implemented citations - Speech recognition - Enabled debugger testing. Memory confirmed. Next: role & streamed responses (streamed gives error). - Docstring added - Perplexity MVP: responds to user - HTTS request based. Streamed responses not used yet. - Added Luma and Perplexity tests ## Key source files (fetched from GitHub, selected and truncated for size) ### main.py ```python from tester import ask_perplexity, insert_perplexity from speechrecog import get_speech from texttosspeech import tts, play_audio, save_audio live_convo = True user_turn = 0 status = ['Not initialized', 'User speaking', 'Tester speaking'] company = 'Big Tech Co.' interview_topic = 'Electrical Engineering' job = 'Junior VLSI designer' # tailor_responses = 'Please also rate on a scale of 1-10 the satisfactoriness of the answer and reply with just the number.' request_summary = 'Please summarize the prior discussion. Tell me how I could have improved my interview answers, then we can move on to behavioral questions.' added_text_prompt = '' n_behavioral = 2 n_end = 5 iteration = 0 def parse_reply(reply): out = ''.join([i for i in reply if (i.isalpha() or i.isspace() or i in [',', '.', '?', '!', "'", "-", ";", "+", "=", "*"])]) return out while (live_convo): # if (iteration == n_rounds): # response = insert_perplexity('Please proceed to the next stage of the interview.') # # response = insert_perplexity('Please begin the behavioral interview stage. Ask questions that reveal how the candidate acts in difficult situations.') # print(response) if (user_turn == 0): user_turn = 1 print('asking perplexity') insert_perplexity("Your role is an interviewer at {}. Please initiate the interview by introducing yourself, and telling the candidate the structure of the interview. Wait for the candidate confirmation before asking technical questions. The topic is {}. The candidate is interviewing for job {}. Do not apologize unnecessarily!".format(company, interview_topic, job)) # reply, citations = response # reply = parse_reply(reply) # print(reply) user_spoken = '' if (user_turn == 1): print(status[user_turn]) user_spoken = get_speech() print(user_spoken) if (user_spoken != ''): user_turn = -1 * user_turn if(user_turn == -1): print(status[user_turn]) response = ask_perplexity(user_spoken + added_text_prompt) # print(response) try: reply, citations = response reply = parse_reply(reply) print(reply) # agent_spoken = tts(reply) # play_audio(agent_spoken) except ValueError: print(response) user_turn = -1 * user_turn if (iteration == n_behavioral): added_text_prompt = request_summary iteration += 1 ``` ### app.py ```python from flask import Flask, render_template, request, url_for, jsonify, redirect, send_from_directory from tester import ask_perplexity, insert_perplexity from texttosspeech import tts, save_audio, play_audio import speech_recognition as sr import os from pydub import AudioSegment import logging # import soundfile # import wave # Initialize the Flask application app = Flask(__name__) app.config['UPLOAD_FOLDER'] = 'static/audio' ALLOWED_EXTENSIONS = {'wav'} AUDIO_DIRECTORY = 'static/audio' def allowed_file(filename): return '.' in filename and \ filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS def convert_audio_to_wav(file_path): """Convert the audio file to a standard PCM WAV format.""" sound = AudioSegment.from_file(file_path) wav_path = file_path.rsplit('.', 1)[0] + '.wav' # Ensure mono channel, 16-bit, and 16kHz sample rate sound = sound.set_channels(1).set_frame_rate(16000).set_sample_width(2) sound.export(wav_path, format='wav') return wav_path def transcribe_audio(file_path): # data, samplerate = soundfile.read(file_path) # new_filename = os.path.join(app.config['UPLOAD_FOLDER'], 'audio.wav') # soundfile.write(new_filename, data, samplerate, subtype='PCM_16') recognizer = sr.Recognizer() recognizer.energy_threshold = 386 # ChatGPT version file_path = convert_audio_to_wav(file_path) with sr.AudioFile(file_path) as source: audio = recognizer.record(source) try: text = recognizer.recognize_google(audio) # with wave.open(file_path, 'rb') as wf: # audio_data = wf.readframes(wf.getnframes()) # audio = sr.AudioData(audio_data, wf.getframerate(), wf.getsampwidth()) # try: # text = recognizer.recognize_google(audio) return text, 0 except sr.UnknownValueError: return "Could not understand audio", 1 except sr.RequestError as e: return f"Could not request results; {e}", 1 def parse_reply(reply): out = ''.join([i for i in reply if (i.isalpha() or i.isspace() or i in [',', '.', '?', '!', "'", "-", ";", "+", "=", "*"])]) return out # Define the home route with a form @app.route('/', methods=['GET', 'POST']) def home(): if request.method == 'POST': name = request.form['name'] message = request.form['message'] return render_template('result.html', name=name, message=message) return render_template('index.html') @app.route('/ask', methods=['POST']) def ask(): if 'file' not in request.files: return jsonify({'response': 'No file part'}), 400 file = request.files['file'] if file.filename == '': return jsonify({'response': 'No selected file'}), 400 if file and allowed_file(file.filename): filename = file.filename file_path = os.path.join(app.config['UPLOAD_FOLDER'], filename) file.save(file_path) # logging.info(f'File saved to {file_path}') status = 1 transcription, status = transcribe_audio(file_path) # logging.info(f'Transcribed text: {transcription}') user_input = transcription reply = None if (status == 0): response = ask_perplexity(transcription) reply, citation = response reply = parse_reply(reply) audio = tts(reply) save_audio(audio, os.path.join(app.config['UPLOAD_FOLDER'], 'reply.mp3')) play_audio(audio) if (status != 0): user_input = 'Please record it again.' return jsonify({'user_input': user_input, 'status': status, 'response': reply, 'audio_filename': os.path.join(app.config['UPLOAD_FOLDER'], 'reply.mp3')}) @app.route('/updateparams', methods=['POST']) def updateparams(): if request.method == 'POST': company = request.form['company'] field = request.form['field'] role = request.form['role'] insert_perplexity("Your name is Connor. Your role is an interviewer at {}. Please initiate the interview by introducing yourself, and telling the candidate the structure of the interview. Wait for the candidate confirmation before asking technical questions. The topic is {}. The candidate is interviewing for job {}. Do not apologize unnecessarily!".format(company, field, role)) print(company, flush=True) return render_template('index.html', company=company, field=field, role=role) return redirect(url_for('home')) # @app.route('/audio/') # def serve_audio(filename): # return send_from_directory(AUDIO_DIRECTORY, filename, as_attachment=True, mimetype='audio/mp3') # Additional route example @app.route('/about') def about(): return render_template('about.html') # The main function to run the app if __name__ == '__main__': app.run(debug=False) # import os # import logging # from flask import Flask, render_template, request, jsonify # from tester import ask_perplexity # import speech_recognition as sr # from pydub import AudioSegment # app = Flask(__name__) # app.config['UPLOAD_FOLDER'] = 'static/audio' # ALLOWED_EXTENSIONS = {'wav', 'mp3'} # def allowed_file(filename): # return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS # def transcribe_audio(file_path): # recognizer = sr.Recognizer() # # Convert MP3 to WAV for compatibility with speech_recognition # if file_path.endswith('.mp3'): # sound = AudioSegment.from_mp3(file_path) # file_path = file_path.replace('.mp3', '.wav') # sound.export(file_path, format='wav') # with sr.AudioFile(file_path) as source: # audio = recognizer.record(source) # try: # text = recognizer.recognize_google(audio) # return text # except sr.UnknownValueError: # return "Could not understand audio" # except sr.RequestError as e: # return f"Could not request results; {e}" # @app.route(' [truncated — 1044 more characters] ``` ### static/js/main.js ```javascript // document.addEventListener("DOMContentLoaded", function() { // const recordButton = document.getElementById('record-btn'); // const responseList = document.getElementById('response-list'); // let mediaRecorder; // let isRecording = false; // let audioChunks = []; // recordButton.addEventListener('click', function() { // if (!isRecording) { // startRecording(); // } else { // stopRecording(); // } // }); // function startRecording() { // navigator.mediaDevices.getUserMedia({ audio: true }) // .then((stream) => { // mediaRecorder = new MediaRecorder(stream); // mediaRecorder.start(); // recordButton.textContent = 'Stop Recording'; // isRecording = true; // audioChunks = []; // mediaRecorder.addEventListener("dataavailable", event => { // audioChunks.push(event.data); // }); // mediaRecorder.addEventListener("stop", () => { // const audioBlob = new Blob(audioChunks, { type: 'audio/wav' }); // convertAndSendToServer(audioBlob); // }); // }) // .catch(error => console.error('Error accessing media devices.', error)); // } // function stopRecording() { // mediaRecorder.stop(); // recordButton.textContent = 'Start Recording'; // isRecording = false; // } // function convertAndSendToServer(audioBlob) { // const reader = new FileReader(); // reader.onload = function(event) { // const arrayBuffer = event.target.result; // const wav = lamejs.WavHeader.readHeader(new DataView(arrayBuffer)); // const samples = new Int16Array(arrayBuffer, wav.dataOffset, wav.dataLen / 2); // const mp3Encoder = new lamejs.Mp3Encoder(1, wav.sampleRate, 128); // const mp3Buffer = []; // let remaining = samples.length; // const maxSamples = 1152; // for (let i = 0; remaining >= maxSamples; i += maxSamples) { // const monoChunk = samples.subarray(i, i + maxSamples); // const mp3buf = mp3Encoder.encodeBuffer(monoChunk); // if (mp3buf.length > 0) { // mp3Buffer.push(new Int8Array(mp3buf)); // } // remaining -= maxSamples; // } // const d = mp3Encoder.flush(); // if (d.length > 0) { // mp3Buffer.push(new Int8Array(d)); // } // const mp3Blob = new Blob(mp3Buffer, { type: 'audio/mpeg-3' }); // sendDataToServer(mp3Blob); // }; // reader.readAsArrayBuffer(audioBlob); // } // function sendDataToServer(audioBlob) { // const formData = new FormData(); // formData.append('file', audioBlob, 'audio.mp3'); // fetch('/ask', { // method: 'POST', // body: formData // }) // .then(response => response.json()) // .then(data => { // const li = document.createElement('li'); // li.textContent = data.response; // responseList.appendChild(li); // }) // .catch((error) => { // console.error('Error:', error); // }); // } // }); document.addEventListener("DOMContentLoaded", function() { const recordButton = document.getElementById('record-btn'); const submitButton = document.getElementById('start-convo'); const responseList = document.getElementById('response-list'); let mediaRecorder; let isRecording = false; recordButton.addEventListener('click', function() { if (!isRecording) { startRecording(); } else { stopRecording(); } }); function startRecording() { navigator.mediaDevices.getUserMedia({ audio: true }) .then((stream) => { mediaRecorder = new MediaRecorder(stream); mediaRecorder.start(); recordButton.textContent = 'Stop Recording'; isRecording = true; const audioChunks = []; mediaRecorder.addEventListener("dataavailable", event => { audioChunks.push(event.data); }); mediaRecorder.addEventListener("stop", () => { const audioBlob = new Blob(audioChunks, { type: 'audio/wav' }); // uncomment to use wav format instead const audioUrl = URL.createObjectURL(audioBlob); const audio = new Audio(audioUrl); // audio.play(); sendDataToServer(audioBlob); // convertAndSendToServer(audioBlob); }); }) .catch(error => console.error('Error accessing media devices.', error)); } function stopRecording() { console.log('recording stopped'); mediaRecorder.stop(); recordButton.textContent = 'Start Recording'; isRecording = false; } // submitButton.addEventListener('submit', function(event) { // event.preventDefault(); // const company = document.getElementById('company').value; // const jobfield = document.getElementById('job-field').value; // const jobrole = document.getElementById('job-role').value; // if (!jobfield || !jobrole) return; // fetch('/updateparams', { // method: 'POST', // headers: { // 'Content-Type': 'application/json' // }, // body: JSON.stringify({ field: jobfield, name: jobrole, company: company }) // }) // .then(response => response.json()) // .then(data => {}) // .catch((error) [truncated — 3934 more characters] ``` ### eltest.py ```python from elevenlabs import generate, set_api_key set_api_key('sk_e5f53a150de75465187a3fa312e6cbca028f3454447133ca') audio = generate( text="Hello World!", voice="Bella" ) with open("out.wav", "wb") as fp: fp.write(audio) ``` ### perplexitytest.py ```python from openai import OpenAI YOUR_API_KEY = "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" messages = [ { "role": "system", "content": ( "You are an artificial intelligence assistant and you need to " "engage in a helpful, detailed, polite conversation with a user." ), }, { "role": "user", "content": ( "How many stars are in the universe?" ), }, ] client = OpenAI(api_key=YOUR_API_KEY, base_url="https://api.perplexity.ai") # chat completion without streaming response = client.chat.completions.create( model="sonar-pro", messages=messages, ) print(response) # chat completion with streaming response_stream = client.chat.completions.create( model="sonar-pro", messages=messages, stream=True, ) for response in response_stream: print(response) ``` ### texttosspeech.py ```python from dotenv import load_dotenv from elevenlabs.client import ElevenLabs from elevenlabs import play, stream, save # ELEVENLABS_API_KEY = 'sk_e5f53a150de75465187a3fa312e6cbca028f3454447133ca' ELEVENLABS_API_KEY = 'sk_a1af276616b5ec2671ea1dfd2f5345b2a1135569160964bf' load_dotenv() client = ElevenLabs(api_key = ELEVENLABS_API_KEY) voiceID = 'EkK5I93UQWFDigLMpZcX' # voiceID = 'TX3LPaxmHKxFdv7VOQHJ' def tts(saytext): audio = client.text_to_speech.convert( text=saytext, voice_id=voiceID, model_id="eleven_multilingual_v2", output_format="mp3_44100_128", ) return audio def save_audio(audio, title="reply.mp3"): save(audio, title) def play_audio(audio): play(audio) # audio_stream = client.text_to_speech.convert_as_stream( # text="This is a test", # voice_id="JBFqnCBsd6RMkjVDRZzb", # model_id="eleven_multilingual_v2" # ) # # option 1: play the streamed audio locally # stream(audio_stream) ``` ### lumatest.py ```python import os import requests import time from lumaai import LumaAI, AsyncLumaAI os.environ["LUMAAI_API_KEY"] = "luma-1c2a2983-47d5-4305-a3e6-7b10596fba7e-3fe1e6aa-bc71-4aa8-8723-bd0215c17360" client = LumaAI( auth_token=os.environ.get("LUMAAI_API_KEY"), ) print(client) generation = client.generations.create( # prompt="Filmed from point of view of a candidate getting interviewed by a manager for a job position. Manager is looking at the camera directly speaking to it. Background is a serious office. Make him say 'hello'.", prompt="A serious character in a suit is talking to the viewer. They are talking and making exaggerated expressions. Their mouth is open sometimes.", ) completed = False while not completed: generation = client.generations.get(id=generation.id) if generation.state == "completed": completed = True elif generation.state == "failed": raise RuntimeError(f"Generation failed: {generation.failure_reason}") print("Dreaming") time.sleep(3) video_url = generation.assets.video # download the video response = requests.get(video_url, stream=True) with open(f'{generation.id}.mp4', 'wb') as file: file.write(response.content) print(f"File downloaded as {generation.id}.mp4") ``` ### speechrecog.py ```python import speech_recognition as sr import pyttsx3 # Initialize the recognizer r = sr.Recognizer() # Function to convert text to # speech def speakText(command): # Initialize the engine engine = pyttsx3.init() engine.say(command) engine.runAndWait() # Loop infinitely for user to # speak def get_speech(): MyText = '' r = sr.Recognizer() r.energy_threshold = 386 # r.dynamic_energy_threshold = False try: with sr.Microphone() as source2: # wait for a second to let the recognizer # adjust the energy threshold based on # the surrounding noise level # TODO: make this adjustable for slower speakers who are more hesitant. r.pause_threshold = 2.0 r.adjust_for_ambient_noise(source2, duration=0.5) #listens for the user's input audio2 = r.listen(source2, timeout=None) # Using google to recognize audio MyText = r.recognize_google(audio2) MyText = MyText.lower() except sr.RequestError as e: print('Could not request results {0}'.format(e)) except sr.UnknownValueError: print('unknown error occurred') return MyText # while(1): # # Exception handling to handle # # exceptions at the runtime # try: # # use the microphone as source for input. # with sr.Microphone() as source2: # # wait for a second to let the recognizer # # adjust the energy threshold based on # # the surrounding noise level # r.adjust_for_ambient_noise(source2, duration=0.2) # #listens for the user's input # audio2 = r.listen(source2) # # Using google to recognize audio # MyText = r.recognize_google(audio2) # MyText = MyText.lower() # print('Did you say', MyText) # SpeakText(MyText) # except sr.RequestError as e: # print('Could not request results {0}'.format(e)) # except sr.UnknownValueError: # print('unknown error occurred') ``` ### tester.py ```python """ Implements the tester back-end using perplexity. Feature Requirements: - Capability of mistake-catching - Capability of suggesting industry standard solutions #TODO: need to confirm - Follow up questions - Design questions - Citations - Memory in the conversation - Fast responses - Streamed responses Features Implemented/Guaranteed: - Output response - Capability of mistake-catching - Citations - Memory - Follow up questions Date: 02/15/25 Author: Andrea Miramontes Serrano """ import requests import json # Andrea's key API_KEY = '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' URL = "https://api.perplexity.ai/chat/completions" CONV_HIST = [ { "role": "system", "content": "You are an interviewer conducting a job interview. Please keep responses short and conversational. To candidate's responses ask follow up, technical questions. If the user makes a factual mistake, ask them about that. Questions are industry-standard. If candidate keeps making mistakes, address their underlying misconception. If the conversation drifts away from the interview topic, do guide the interviewee back." # You are an interviewer conducting a job interview. # Keep responses short and conversational. # Catch mistakes, provide sources and always ask a follow up, technical question. # If the user makes a mistake, ask them about that. Every once in a while, ask design questions and/or industry-standard questions. # If there are mistakes that keep happening, detect when the user has an underlying misconception. # """ } ] def insert_perplexity(admin_prompt: str): CONV_HIST.append({"role": "system", "content": admin_prompt}) # initial_content = CONV_HIST[0]["content"] # output_content = initial_content + admin_prompt # CONV_HIST[0]["content"] = output_content # response = requests.post(URL, headers=headers, data=json.dumps(data)) # if response.status_code == 200: # result = response.json() # text_response, citations = result["choices"][0]["message"]["content"], result["citations"] # # Add AI response to memory # CONV_HIST.append({"role": "assistant", "content": text_response}) # return text_response, citations # else: # return f"Error: {response.status_code}, {response.text}" def ask_perplexity(prompt: str, role:str = "user") -> str: """ Initializes a perplexity API call: - Asks for interview context #TODO: expand feature set to also social anxiety context - enables user to choose prompt #TODO: expand to have this be a voice input. - Capability of mistake-catching - Capability of suggesting industry standard solutions #TODO: need to confirm - Follow up questions - Design questions - Citations - Memory in the conversation - Fast responses - Streamed responses """ headers = { "Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json" } CONV_HIST.append({"role": role, "content": prompt}) data = { "model": "sonar-pro",# Use best available model "messages": CONV_HIST, # Maintain conversation memory "max_tokens": 300, "temperature": 0.7, "top_p": 0.9, "top_k": 0, "stream": False, "presence_penalty": 0, "frequency_penalty": 0.1 # Set to a valid value (>0) } response = requests.post(URL, headers=headers, data=json.dumps(data)) if response.status_code == 200: result = response.json() text_response, citations = result["choices"][0]["message"]["content"], result.get("citations") # Add AI response to memory CONV_HIST.append({"role": "assistant", "content": text_response}) return text_response, citations else: return f"Error: {response.status_code}, {response.text}" #TODO: clean up for streamed responses - this bit chunks the response. # full_response = "" # print("\n📢 AI is responding...\n") # for line in response.iter_lines(): # if line: # decoded_line = json.loads(line.decode("utf-8")) # text_chunk = decoded_line["choices"][0]["message"]["content"] # print(text_chunk, end="", flush=True) # Show response progressively # full_response += text_chunk # Another way of extracting citations [doesn't work as well]: # references = decoded_line["choices"][0]["message"].get("references", []) # if references: # print("\n\n🔗 **Citations:**") # for ref in references: # print(f"- {ref['title']} ({ref['url']})") # ----------------- USAGE ----------------- # import sys #TODO: remove - only here for debugging purposes # if len(sys.argv) > 1: # user_input = sys.argv[1] # Takes input from launch.json # else: # user_input = input("Enter your prompt: ") # Fallback manual input # company = 'Big Tech Co.' # interview_topic = 'Electrical Engineering' # job = 'Junior VLSI designer.' # n_rounds = 2 # iteration = 0 # insert_perplexity("Your role is an interviewer at {}. Please initiate the interview by introducing yourself, and telling the candidate the structure of the interview. Wait for the candidate confirmation before asking technical questions. The topic is {}. The candidate is interviewing for job {}. Do not apologize unnecessarily! After {} rounds of technical questions, move on to behavioral questions in the interview.".format(company, interview_topic, job, n_rounds)) # print(CONV_HIST[0]) # while (True): # response = ask_perplexity(user_input) # ask_perplexity(user_input) # print("\nPerplexity AI Response:\n", response[0]) # user_input = input("Enter your prompt: ") ``` ### templates/about.html ```html About Page

About Us

This is the about page of the web app.

Home ``` [2 more indexed source files omitted to keep this export small. The full file list is in the Codebase structure section above.]