ConsultNow

• 

  Notification notifying a customer after sending an appointment request to the Expert

• 

  A "Free One hour" consultation appointment request have been notified to the Expert

• 

  A notification email notifying Customer after an Expert have accepting the meeting request

• 

  A notification notifying to Customer after an Expert rejecting Consultation request

• 

  A meeting invitation link sent to Customer after an Expert accepting the customer request

• 

  An expert feedback notification request shared to Customer after an Expert consultation

• 

  An email follow up by expert to the Customer using an AI Session follow up feature in the Expert dashboard

• 

  An alert notification sent from the Consultnow admin team to the Expert After He enroll's as an Expert in the Consultnow application

• 

  A profile evaluation/verification from Consultnow team to the newly enrolled experts in the Consultnow application

• 

  An expert dashboard profile widget allowing an expert to prepare for the Customer query

• 

  A post meeting synopsys notification to the expert, consultant and the admin team

• 

  An AI post session follow up email generator from the Expert dashboard

• 

  A gemini 2.5 flash engine use in the Consultnow application

• 

  A flow diagram demonstrating journey of the customer and an AI triage in the Consultnow application

• 

  A complete razor payment workflow in the Consultnow application

Inspiration

The core inspiration for this project stems from my family’s passion for helping others through their diverse expertise. My mother, a former schoolteacher, wanted to utilize her free time to teach Telugu and Mathematics to children both locally and abroad. Initially, my thought was to simply string together WhatsApp meeting invites and manual payment links. However, the XPRIZE Hackathon presented the perfect opportunity to build her a dedicated, scalable platform.

As I began conceptualizing the project using Spec-Driven Development and Gemini 2.5 Flash in VS Code, the vision expanded. During a conversation with my fiancée—who excitedly joined me as a teammate—and my father, who wanted to start a legal service company, the idea evolved into a multi-vertical consultation marketplace. With my sibling and fiancée both being doctors, we initially grouped our expert categories into "Medical Advice," "Legal Advice," "Student Tutoring," "IT Career Guidance," and "HR Services."

While navigating real-world business compliance, our payment gateway (Razorpay) advised us against highly regulated sectors like medical and legal advice. We swiftly pivoted our application to focus on the high-value, lower-risk categories of Tutoring, IT, and HR to secure live payment access. Ultimately, the overwhelming viewership and support from my family served as the driving inspiration to build this enterprise-level solution.

What it does

ConsultNow is an AI-Native, on-demand professional services marketplace that connects users with vetted experts for 1-on-1 live video consultations. It transforms the traditional directory model into a smart, frictionless ecosystem.

• AI-Powered Triage & Support: Users don't just blindly search for help; our integrated AI assists them in formulating their queries correctly and routes their specific problem to the perfect expert category.
• Interactive AI Expert Twins: Clients can chat with roleplay AI avatars of experts before booking a live session, helping clarify objectives and build booking confidence.
• Frictionless Booking & Scheduling: Users can browse expert profiles, select a time slot, apply promo codes, and securely checkout. The platform automatically provisions video links and sends calendar invites to both parties.
• "1st Hour Free" Promotional Hook: To lower the barrier to entry, build trust, and drive user acquisition, the platform allows users to claim a 1-hour free consultation with verified experts.
• Automated Marketing & Pricing for Experts: Upon registration, the platform’s AI automatically generates professional, SEO-optimized marketing bios, recommends hourly pricing based on market benchmarks, and builds embeddable website widgets.
• Post-Consultation AI Intelligence: Automatically generates 30-second intake briefings for experts before calls, creates post-session HTML synopsis reports with Session Effectiveness Scores (e.g. 9.5/10), and extracts sentiment word bubbles from client reviews.

How we built it

We architected a modern, full-stack Single Page Application (SPA) leveraging AI not just as a coding assistant, but as a core operational feature of the platform. We utilized Spec-Driven Development with VS Code, Gemini IDE agents, and Anti Gravity Studio to ensure zero-hallucination code generation and rapid UI/UX iteration.

Technical Stack & Architecture

Detailed AI Pipeline Architecture (11 Gemini 2.5 Flash Touchpoints)

1. Natural Language Problem Triage (POST /api/ai/triage): Classifies user issue descriptions into Tutoring, IT Career, or HR Services with reasoning.
2. Interactive AI Twin Roleplay (POST /api/ai/expert-twin-chat): Roleplays expert avatars using bio and credentials to answer pre-booking client questions.
3. Smart Recommendation Badges (POST /api/ai/expert-summaries): Generates 1-sentence personalized match badges rendered on expert search cards.
4. Bio Copywriting & Marketing Generator (POST /api/ai/generate-marketing): Converts expert skills into professional bios and high-converting taglines.
5. Hourly Pricing Advisor (POST /api/ai/recommend-pricing): Benchmarks years of experience and domain demand to suggest optimal INR rates.
6. Pre-Session Client Prep Agenda (POST /api/ai/agenda): Produces a 3-step preparation checklist (documents, questions, goal alignment) for clients before meetings.
7. 30-Second Pre-Call Intake Briefing (POST /api/ai/briefing): Synthesizes client booking notes into a structured intake digest for expert dashboards.
8. Post-Call Follow-Up & Action Plan (POST /api/ai/followup): Drafts personalized follow-up emails and task lists for experts to send post-session.
9. Programmatic SEO Generation (POST /api/seo/generate/:expertId): Auto-generates 300-word SEO bios, targeted service offerings, FAQs, and OpenGraph tags (/expert/seo/:slug).
10. Automated Meeting Synopsis & Evaluation Score (POST /api/bookings/:id/synopsis): Generates HTML meeting digests with Session Effectiveness Scores (e.g., 9.5/10) emailed to experts.
11. Feedback Sentiment Word Bubbles (POST /api/feedback): Extracts 3-6 descriptor tags (e.g., "Insightful", "Punctual") from client reviews to render word bubbles on profile cards.

Challenges we ran into

While AI-assisted development drastically accelerated our coding phase, taking the application to a production-ready business introduced significant real-world challenges:

• Regulatory Compliance & Payments: Our initial inclusion of "Medical" and "Legal" categories led to a Razorpay KYC rejection due to strict Indian regulations. We had to swiftly refactor the application's UI, routing, and database schema to sanitize the platform, successfully appealing the rejection by pivoting to Tutoring, IT, and HR.
• API Rate Limiting & Response Latency: During high-traffic generation of expert marketing bios and triage queries, we encountered Gemini API rate limits (Error 429) and formatting delays. We solved this by engineering an exponential backoff retry handler (generateWithRetry) in Node.js and configuring Gemini 2.5 Flash with native JSON mode (responseMimeType: "application/json").
• Customer Acquisition & Trust: Moving from a deployed app to acquiring real users is tough. We had to quickly learn digital marketing, deploying tightly fenced geographic ad campaigns across Google Ads, Facebook, Instagram, Threads, and X while offering 1-hour free promotional consultations to establish early trust.

Accomplishments that we're proud of

• End-to-End Execution: Successfully designing, developing, and deploying a scalable, full-stack SPA with complex third-party API integrations (Razorpay Payments, Socket.io WebSockets, Nodemailer SMTP, and Gemini 2.5 Flash LLMs) within the hackathon timeframe.
• Real-World Traction: Transitioning from a purely technical project to a live business. Our multi-channel marketing campaigns successfully attracted real users, resulting in actual experts registering on the platform and real customers booking the 1-hour free consultation.
• AI-Native Operations: We are incredibly proud of pushing beyond standard CRUD functionality by embedding Gemini directly across 11 touchpoints in the business logic to automate operational overhead (triage, AI twins, marketing, intake briefs, and post-meeting synopsis reports).

What we learned

• Mastering AI-Assisted Engineering: We learned how to effectively write strict architectural markdown files (spec.md and gemini.md) to guide autonomous IDE agents like Anti Gravity and VS Code without breaking existing logic.
• Business & Compliance: We gained invaluable firsthand experience navigating fintech compliance, payment gateway regulations, and the difference between building a prototype and a legal, monetizable business.
• Digital Marketing Mechanics: We learned how to deploy long-tail keywords, optimize ad spend using geo-targeting, and leverage our platform's AI to generate localized marketing copy for social media.

What's next for ConsultNow

Our immediate goal is to scale our active user base and secure a broader network of global experts. To support this growth, our technical roadmap includes:

• Database & API Scaling: Expanding Redis caching to improve load times and scaling our backend infrastructure to handle high-concurrency WebSocket connections for live video/chat.
• Post-Consultation AI Transcript Summaries: Using Gemini to ingest live chat/video transcripts to automatically generate personalized action plans and transcript highlights for users after their consultation ends.
• "Trust & Verifiability" Dashboard: Building out an aggregation metric dashboard that displays an expert's average response time, free-tier acceptance rate, and successful session completions to build deep trust with our users.

Built With

• angular.js
• antigravity
• gcp
• gemini
• geminiflash3.5
• google-calendar
• googlemeet
• netlify
• node-mailer
• node.js
• razorpay
• render
• upstash