No-Code AI-Based Interview Prep Tool
Budget: $1,500 – $3,000 USD
Title: Build a No-Code AI Interview Prep MVP (Bubble / OpenAI API)
Looking for a fast, experienced no-code developer to build a text-based executive interview prep tool. Core features: User login, resume/job description upload, an interactive text-based chat interface connected to OpenAI API with strict persona prompts, a basic scoring/feedback screen, and standard Stripe payment gating for a 90-day pass. No complex database architecture required. Speed to market is priority.
STATEMENT OF WORK (SOW) & PROJECT DESCRIPTION
PROJECT NAME: Catalyst Executive Interview Coach (Phase 1 MVP)
Target Build Timeline: 30–45 Days
Target Stack: Bubble.io OR Lovable + Xano (Backend) + Stripe API + OpenAI (Whisper & GPT-4o APIs)
1. PROJECT OVERVIEW & OBJECTIVE
The objective of this project is to build a high-performance, minimalist web application that acts as an "anti-validation" interview simulator for executive-level candidates (C-Suite, VP, Directors).
The application will ingest a user's resume, a target job description, and chosen parameters to generate highly realistic, high-pressure interview questions. The core user interaction requires the user to verbally answer questions via an audio-recording mechanism. The system will transcribe the audio via API, analyze the text against strict corporate communication rubrics, return data-driven forensic scores, and generate a optimized "Executive Rewrite" script.
2. SYSTEM ARCHITECTURE & DATA SCHEMAS
To keep development rapid and clean, the database must follow a strict Parent-Child relational hierarchy.
A. Data Table 1: Company_Pipelines (Parent Table)
Tracks the overarching job applications the user is preparing for.
Pipeline_ID (Unique String, Primary Key)
User_ID (Linked to User Account)
Company_Name (Text)
Job_Title (Text)
Job_Description_Raw (Long Text)
Resume_Data_URL (File Storage Link for PDF/Doc)
Created_Timestamp (Date/Time)
B. Data Table 2: Practice_Sessions (Child Table)
Tracks every individual multi-question simulation run executed under a specific company pipeline.
Session_ID (Unique String, Primary Key)
Pipeline_ID (Foreign Key linking to Table 1)
Timestamp (Date/Time)
Persona_Type (Text: Recruiter, Hiring Manager, CEO, Board Member)
Interview_Stage (Text: Panel, Board Review, etc.)
Overall_Session_Score (Number / Percentage)
C. Data Table 3: Question_Logs (Grandchild Table)
Tracks every individual question asked within a single practice session, storing the transcripts and AI payloads.
Question_ID (Unique String, Primary Key)
Session_ID (Foreign Key linking to Table 2)
Question_Text (Text)
Raw_Spoken_Transcript (Long Text)
Score_Clarity (Number 1-10)
Score_Presence (Number 1-10)
Score_Strategy (Number 1-10)
Score_Conciseness (Number 1-10)
AI_Critique_Text (Long Text)
AI_Executive_Rewrite (Long Text)
3. CORE INTERFACE LAYOUT & UI REQUIREMENTS
The interface must be clean, minimalist, and high-contrast, avoiding unnecessary animations or decorative elements.
+-----------------------------------------------------------------------+
| [Logo] Catalyst Executive Coach [Dashboard] [My Account]|
+-----------------------------------------------------------------------+
| INTERVIEWER PANEL | CURRENT QUESTION DISPLAY |
| | "Walk me through your 90-day turnaround..." |
| Name: Marcus Vance | |
| Title: Skeptical CEO| +-------------------------------------------+ |
| Style: Ruthless | | [ Click to Record Answer] | |
| | +-------------------------------------------+ |
| | | Transcript preview box populates here... | |
| | +-------------------------------------------+ |
| | [Submit Answer] [Skip] [Request Hint] |
| |-------------------------------------------------|
| | FORENSIC ANALYTICS (Populates after submission)|
| | Clarity: 6/10 | Presence: 5/10 | Strategy: 8/10|
| | Red Flags: Rambled on micro-mechanics. |
| | Executive Rewrite: [Hand them the script] |
+-----------------------------------------------------------------------+
Screen 1: Executive Dashboard
Displays a grid of user-created Company_Pipelines cards (e.g., "CRO Interview at Stripe").
Each card must show: Job Title, Company Name, Total Practice Runs Executed, Initial Practice Score, and Highest Peak Score reached.
Clicking a card opens the Nested Historical Log, showing a chronological list of past Practice_Sessions with direct links to view previous answers and scripts.
Button: [+ Create New Target Interview Pipeline]
Screen 2: Setup & Configuration Intake
Text input fields for Company Name and Job Title.
File uploader block for Resume (PDF/Doc parsing ready).
Paste box for Job Description.
Selectable Dropdowns for Interview Stage and Interviewer Persona Matrix (Restricted based on Stripe Subscription Tier).
Screen 3: The Interaction Workspace
Left Column: Presenting the Interviewer Profile Card (Name, Title, Style).
Center Column Top: The generated text question displayed in prominent typography.
Center Column Middle: The Audio Interaction Block. A clear mic toggle button. While recording, show a simple recording state. On stop, it instantly updates the editable text transcript box below it.
Center Column Bottom: Action utility row containing buttons: [Submit Answer], [Skip Question], [Request Strategic Hint], [End Interview].
Right Column (Hidden until [Submit Answer] is fired): Reveals the analytics panel displaying the 1-10 sliders, bulleted red flags, and the copyable text block containing the optimized Executive Rewrite.
4. SCOPE OF WORK (SOW) & DELIVERABLES BY PHASE
The developer must build and deliver Phase 1 completely before moving forward.
Phase 1: Core Functional MVP (The 30-Day Sprint)
Deliverable 1.1: Responsive UI/UX Shell Built cleanly in Bubble/Lovable following a strict, dark/minimalist theme layout.
Deliverable 1.2: Authentication & Stripe Gateways Simple registration/login. Setup of Stripe checkout redirection handling three states: Free Trial User, 7-Day Cram Pass User, 90-Day Executive Pass User.
Deliverable 1.3: Low-Cost Verbal Intake Flow Integration with OpenAI Whisper API. When a user clicks "Record Answer," the app captures raw audio via browser microphone access, ships the audio file to Whisper, and pipes the returning text string into the user response box.
Deliverable 1.4: Multi-Persona LLM Prompt Routing Engine Wiring up the OpenAI GPT-4o API. The backend must ingest the Parent Pipeline data (Resume + Job Desc) and pass it into 4 distinct system prompts based on the selected persona (Friendly Recruiter, Hiring Manager, Skeptical CEO, Board Member).
Deliverable 1.5: The Analytics Compiler & Rewrite Engine Logic to split the LLM response payload securely into three UI components: 1-10 integers for metrics, an array of strings for Red Flags, and a markdown text block for the complete Executive Rewrite script.
Deliverable 1.6: Historical Log Database Architecture Ensuring all metrics, transcripts, and rewrites are nested permanently under that specific company card for sequential review.
Out of Scope for Phase 1 (Do Not Build)
Native voice synthesis (AI speaking out loud).
Video avatars / synthetic streaming faces.
Chrome Extensions or automated job board scrapers.
5. REVENUE MECHANICS & TIER LOGIC
(STRIPE INTEGRATION)
The application developer must configure permissions exactly according to the user's active billing token:
Free Tier Token Rules:
Restrict database entries to maximum 1 Company Pipeline.
Restrict total interview questions to 3.
Lock out all personas except The Friendly Recruiter.
Hide/Blur the AI_Executive_Rewrite block post-submission, replacing it with a custom Stripe checkout CTA.
Tier 1: Emergency Cram Pass ($29.99 Flat Fee for 7 Days):
Access duration automatically terminates via backend workflow exactly 168 hours post-purchase.
Access restricted to maximum 2 Company Pipelines, 5 questions per session.
Lock out Skeptical CEO and Board Member personas. Full visibility on text scores and rewrites.
Tier 2: Executive Sprint Pass ($99.00 Flat Fee for 90 Days):
Access duration terminates exactly 90 days post-purchase.
Unlimited Pipelines, Unlimited Questions, Unlimited History Logging.
All 4 core personas completely unlocked. Full activation of the blunt, anti-validation critique prompts.
6. TECHNICAL ACCEPTANCE CRITERIA FOR PAYMENTS
Milestone payouts to the freelancer will only be released once the following functional loops are verified:
The Audio Latency Test: Speech-to-text conversion (Record -> Whisper -> UI Transcript Box text population) must take fewer than 3.5 seconds for a standard 30-second spoken response.
The Database Separation Test: Logging into User Account A must never leak or display historical pipeline cards, scorecards, or uploaded resumes belonging to User Account B.
The Paywall Hard Fence: Manually manipulating the browser URL pathing to access /workspace while holding a Free Tier flag must automatically force-redirect the session back to the checkout portal if parameters exceed 3 questions.
Prompt Format Rigidity: The LLM output must reliably parse into structured frontend fields without breaking layout boxes or dumping raw JSON markers onto the screen.
Looking for a fast, experienced no-code developer to build a text-based executive interview prep tool. Core features: User login, resume/job description upload, an interactive text-based chat interface connected to OpenAI API with strict persona prompts, a basic scoring/feedback screen, and standard Stripe payment gating for a 90-day pass. No complex database architecture required. Speed to market is priority.
STATEMENT OF WORK (SOW) & PROJECT DESCRIPTION
PROJECT NAME: Catalyst Executive Interview Coach (Phase 1 MVP)
Target Build Timeline: 30–45 Days
Target Stack: Bubble.io OR Lovable + Xano (Backend) + Stripe API + OpenAI (Whisper & GPT-4o APIs)
1. PROJECT OVERVIEW & OBJECTIVE
The objective of this project is to build a high-performance, minimalist web application that acts as an "anti-validation" interview simulator for executive-level candidates (C-Suite, VP, Directors).
The application will ingest a user's resume, a target job description, and chosen parameters to generate highly realistic, high-pressure interview questions. The core user interaction requires the user to verbally answer questions via an audio-recording mechanism. The system will transcribe the audio via API, analyze the text against strict corporate communication rubrics, return data-driven forensic scores, and generate a optimized "Executive Rewrite" script.
2. SYSTEM ARCHITECTURE & DATA SCHEMAS
To keep development rapid and clean, the database must follow a strict Parent-Child relational hierarchy.
A. Data Table 1: Company_Pipelines (Parent Table)
Tracks the overarching job applications the user is preparing for.
Pipeline_ID (Unique String, Primary Key)
User_ID (Linked to User Account)
Company_Name (Text)
Job_Title (Text)
Job_Description_Raw (Long Text)
Resume_Data_URL (File Storage Link for PDF/Doc)
Created_Timestamp (Date/Time)
B. Data Table 2: Practice_Sessions (Child Table)
Tracks every individual multi-question simulation run executed under a specific company pipeline.
Session_ID (Unique String, Primary Key)
Pipeline_ID (Foreign Key linking to Table 1)
Timestamp (Date/Time)
Persona_Type (Text: Recruiter, Hiring Manager, CEO, Board Member)
Interview_Stage (Text: Panel, Board Review, etc.)
Overall_Session_Score (Number / Percentage)
C. Data Table 3: Question_Logs (Grandchild Table)
Tracks every individual question asked within a single practice session, storing the transcripts and AI payloads.
Question_ID (Unique String, Primary Key)
Session_ID (Foreign Key linking to Table 2)
Question_Text (Text)
Raw_Spoken_Transcript (Long Text)
Score_Clarity (Number 1-10)
Score_Presence (Number 1-10)
Score_Strategy (Number 1-10)
Score_Conciseness (Number 1-10)
AI_Critique_Text (Long Text)
AI_Executive_Rewrite (Long Text)
3. CORE INTERFACE LAYOUT & UI REQUIREMENTS
The interface must be clean, minimalist, and high-contrast, avoiding unnecessary animations or decorative elements.
+-----------------------------------------------------------------------+
| [Logo] Catalyst Executive Coach [Dashboard] [My Account]|
+-----------------------------------------------------------------------+
| INTERVIEWER PANEL | CURRENT QUESTION DISPLAY |
| | "Walk me through your 90-day turnaround..." |
| Name: Marcus Vance | |
| Title: Skeptical CEO| +-------------------------------------------+ |
| Style: Ruthless | | [ Click to Record Answer] | |
| | +-------------------------------------------+ |
| | | Transcript preview box populates here... | |
| | +-------------------------------------------+ |
| | [Submit Answer] [Skip] [Request Hint] |
| |-------------------------------------------------|
| | FORENSIC ANALYTICS (Populates after submission)|
| | Clarity: 6/10 | Presence: 5/10 | Strategy: 8/10|
| | Red Flags: Rambled on micro-mechanics. |
| | Executive Rewrite: [Hand them the script] |
+-----------------------------------------------------------------------+
Screen 1: Executive Dashboard
Displays a grid of user-created Company_Pipelines cards (e.g., "CRO Interview at Stripe").
Each card must show: Job Title, Company Name, Total Practice Runs Executed, Initial Practice Score, and Highest Peak Score reached.
Clicking a card opens the Nested Historical Log, showing a chronological list of past Practice_Sessions with direct links to view previous answers and scripts.
Button: [+ Create New Target Interview Pipeline]
Screen 2: Setup & Configuration Intake
Text input fields for Company Name and Job Title.
File uploader block for Resume (PDF/Doc parsing ready).
Paste box for Job Description.
Selectable Dropdowns for Interview Stage and Interviewer Persona Matrix (Restricted based on Stripe Subscription Tier).
Screen 3: The Interaction Workspace
Left Column: Presenting the Interviewer Profile Card (Name, Title, Style).
Center Column Top: The generated text question displayed in prominent typography.
Center Column Middle: The Audio Interaction Block. A clear mic toggle button. While recording, show a simple recording state. On stop, it instantly updates the editable text transcript box below it.
Center Column Bottom: Action utility row containing buttons: [Submit Answer], [Skip Question], [Request Strategic Hint], [End Interview].
Right Column (Hidden until [Submit Answer] is fired): Reveals the analytics panel displaying the 1-10 sliders, bulleted red flags, and the copyable text block containing the optimized Executive Rewrite.
4. SCOPE OF WORK (SOW) & DELIVERABLES BY PHASE
The developer must build and deliver Phase 1 completely before moving forward.
Phase 1: Core Functional MVP (The 30-Day Sprint)
Deliverable 1.1: Responsive UI/UX Shell Built cleanly in Bubble/Lovable following a strict, dark/minimalist theme layout.
Deliverable 1.2: Authentication & Stripe Gateways Simple registration/login. Setup of Stripe checkout redirection handling three states: Free Trial User, 7-Day Cram Pass User, 90-Day Executive Pass User.
Deliverable 1.3: Low-Cost Verbal Intake Flow Integration with OpenAI Whisper API. When a user clicks "Record Answer," the app captures raw audio via browser microphone access, ships the audio file to Whisper, and pipes the returning text string into the user response box.
Deliverable 1.4: Multi-Persona LLM Prompt Routing Engine Wiring up the OpenAI GPT-4o API. The backend must ingest the Parent Pipeline data (Resume + Job Desc) and pass it into 4 distinct system prompts based on the selected persona (Friendly Recruiter, Hiring Manager, Skeptical CEO, Board Member).
Deliverable 1.5: The Analytics Compiler & Rewrite Engine Logic to split the LLM response payload securely into three UI components: 1-10 integers for metrics, an array of strings for Red Flags, and a markdown text block for the complete Executive Rewrite script.
Deliverable 1.6: Historical Log Database Architecture Ensuring all metrics, transcripts, and rewrites are nested permanently under that specific company card for sequential review.
Out of Scope for Phase 1 (Do Not Build)
Native voice synthesis (AI speaking out loud).
Video avatars / synthetic streaming faces.
Chrome Extensions or automated job board scrapers.
5. REVENUE MECHANICS & TIER LOGIC
(STRIPE INTEGRATION)
The application developer must configure permissions exactly according to the user's active billing token:
Free Tier Token Rules:
Restrict database entries to maximum 1 Company Pipeline.
Restrict total interview questions to 3.
Lock out all personas except The Friendly Recruiter.
Hide/Blur the AI_Executive_Rewrite block post-submission, replacing it with a custom Stripe checkout CTA.
Tier 1: Emergency Cram Pass ($29.99 Flat Fee for 7 Days):
Access duration automatically terminates via backend workflow exactly 168 hours post-purchase.
Access restricted to maximum 2 Company Pipelines, 5 questions per session.
Lock out Skeptical CEO and Board Member personas. Full visibility on text scores and rewrites.
Tier 2: Executive Sprint Pass ($99.00 Flat Fee for 90 Days):
Access duration terminates exactly 90 days post-purchase.
Unlimited Pipelines, Unlimited Questions, Unlimited History Logging.
All 4 core personas completely unlocked. Full activation of the blunt, anti-validation critique prompts.
6. TECHNICAL ACCEPTANCE CRITERIA FOR PAYMENTS
Milestone payouts to the freelancer will only be released once the following functional loops are verified:
The Audio Latency Test: Speech-to-text conversion (Record -> Whisper -> UI Transcript Box text population) must take fewer than 3.5 seconds for a standard 30-second spoken response.
The Database Separation Test: Logging into User Account A must never leak or display historical pipeline cards, scorecards, or uploaded resumes belonging to User Account B.
The Paywall Hard Fence: Manually manipulating the browser URL pathing to access /workspace while holding a Free Tier flag must automatically force-redirect the session back to the checkout portal if parameters exceed 3 questions.
Prompt Format Rigidity: The LLM output must reliably parse into structured frontend fields without breaking layout boxes or dumping raw JSON markers onto the screen.