NegotiateAI

• 

  History

• 

  Negotiation Score

• 

  Dasgboard

• 

  Salary Negotiation

• 

  landing

🧠 Inspiration Did you know that failing to negotiate a starting salary can cost you over $1,000,000 over a lifetime? The inspiration for NegotiateAI came from a simple observation: people lose thousands of dollars every year—not because they lack talent, but because they freeze during high-stakes conversations. They accept lowball salary offers, fall for car dealer EMI traps, and pay inflated medical bills. Existing solutions like books and videos offer passive learning, but negotiation is an active skill. There was no safe, realistic "gym" to practice these conversations without real-world risk.

💡 What it does NegotiateAI is a personal AI negotiation simulator. Users can practice realistic conversations against dynamic AI opponents across 7 real-world scenarios:

💰 Salary Negotiation 🚗 Car Purchase 🏠 Rent Discussion 💼 Freelance Rate 🏥 Medical Bill 🤝 Business Deal 💳 Credit Card APR The app provides Quick Tips before each round. The AI opponent sets realistic traps (like distracting users with monthly payments instead of total price). After the negotiation, the app generates a Scorecard scoring the user on Confidence, Strategy, Empathy, and Outcome. Finally, an AI Coach provides a personalized debrief on how to improve.

Frictionless & Collaborative: To ensure zero barriers to practice, NegotiateAI is completely login-free. Instead of a private history page, we feature a 'Community Feed'—allowing users to learn not just from their own AI Coach feedback, but by observing the negotiation strategies and outcomes of the broader community.

🛠️ How I built it I built the entire full-stack application with zero code, entirely on MeDo.

Iterative Generation: I used MeDo's multi-turn chat to build the app step-by-step. First, I prompted the UI and database schema. Then, I added the scenario logic and chat interfaces. Prompt Engineering for AI Opponents: I used MeDo to inject hidden system prompts to make the AI act like a tough HR manager, a pushy car dealer, or a strict billing rep. The Evaluation Engine: The most impressive feature is the scoring system. I leveraged MeDo's AI capabilities to create a secondary prompt that takes the chat transcript, analyzes the conversation patterns, and outputs a structured JSON object to render the visual scorecard and coach feedback. 🧗 Challenges I ran into The biggest challenge was getting the AI opponents to behave realistically. Initially, the AI was too easy and gave in immediately, which didn't help the user learn. I had to iterate my prompts in MeDo, explicitly instructing the AI to use real-world negotiation tactics (like the EMI trap in car sales, or citing "company policy" in salary talks) and only yield if the user used strong counter-tactics. Structuring the AI Coach output to consistently return parseable data for the scorecard was also a fun challenge solved purely through prompt refinement.

🏆 Accomplishments that I'm proud of I am incredibly proud of building a dual-AI system without writing a single line of code. Creating an app where one AI acts as a resistant opponent and a second AI analyzes the conversation to provide structured, visual feedback is something I didn't think was possible without heavy backend coding. I'm also proud of designing a frictionless, login-free experience—by turning the history page into a 'Community Feed', users can learn from each other's negotiation strategies in real-time, creating a collaborative learning environment rather than an isolated one.

🎓 What I learned I learned that the key to building powerful no-code AI applications isn't just about writing a single magic prompt—it's about conversation design. Breaking a complex app down into iterative multi-turn chats on MeDo yields far better results than trying to build everything at once. I also learned how deeply AI can simulate human behavior and provide actionable, personalized feedback when guided correctly.

🚀 What's next for NegotiateAI The journey doesn't stop here! For the next version of NegotiateAI, I plan to:

Voice Negotiation Mode: Integrate speech-to-text and text-to-speech so users can practice their tone of voice and pacing, not just their words. Live Market Data Integration: Connect real-time APIs (like Glassdoor for salaries or Zillow for rent) so the AI opponent uses accurate, up-to-the-minute market data during negotiations. Gamification & Leaderboards: Add badges, streaks, and a community leaderboard to make practice addictive and fun. Corporate Training Modules: Create B2B packages so companies can use NegotiateAI to train their sales and HR teams in-house.

Built With

• ai
• llm
• medo
• no-code
• prompt
• voice-over