DineConnect

A revolutionary engagement-centric micro-influencer restaurant marketing platform designed for Indian market it that connects them with restaurants through a cashback-driven marketing system.

Comment

DineConnect

Inspiration

DineConnect was born from witnessing how micro-influencers drive real customer engagement for restaurants, yet lack transparent, trackable systems to prove ROI. Combining India’s rapid UPI adoption with the explosion of Instagram content creation, we set out to build a platform that rewards genuine engagement—beyond vanity metrics—and delivers measurable value to influencers, restaurants, and diners.

What it does

DineConnect is a performance-driven micro-influencer restaurant marketing ecosystem.

  • Influencers sign up with Instagram OAuth and are automatically assigned an engagement-based tier (Bronze–Platinum) in INR, based on follower count, composite engagement score, and purchase history.
  • Restaurants subscribe in monthly investment tiers (Starter, Growth, Premium) and receive a single QR code per outlet that, when scanned, displays all active influencer codes and base cashback offers.
  • Diners scan the QR, browse partnered influencers and offers, transact via UPI or card, and receive instant cashback, with influencers and the platform receiving simultaneous payouts.
  • A robust RBAC system secures four roles—Platform Admin, Restaurant, Influencer, Follower—with granular, performance-based permissions.
  • Real-time analytics dashboards provide influencer ROI, restaurant customer acquisition metrics, and platform-level insights.

How we built it

Frontend

  • Next.js (v14) + TypeScript with PWA support for SSR/ISR, offline usage, push notifications
  • Tailwind CSS + Radix UI for responsive, accessible components
  • Zustand for lightweight state management
  • React Query for server state caching

Backend Microservices (Golang)

  • Gin framework for HTTP routing
  • PostgreSQL (ACID transactions), Redis (caching & sessions), MongoDB (analytics logs), ClickHouse (time-series metrics)
  • RabbitMQ for event-driven processing (QR scan, transaction, analytics)
  • gRPC for inter-service communication
  • Kubernetes on AWS/GCP with Istio service mesh, Prometheus/Grafana monitoring, and ELK logging

Payments & Cashback

  • Razorpay/Cashfree UPI integration, Stripe for cards
  • Custom fee-structure logic for UPI (0–0.5%), cards (0.5–1.8%), wallets (1.1%)
  • Instant multi-party payouts via event-driven PaymentService

Instagram & Engagement Scoring

  • Instagram Basic Display API for follower and engagement data
  • Composite engagement algorithm weighting likes, comments, shares, saves, video/story views, content quality, audience authenticity
  • Automated tier adjustment via cron jobs and ML-driven performance thresholds

Challenges we ran into

  • Attribution Modeling: Designing multi-touch, time-decay attribution to fairly credit multiple influencers per customer journey.
  • Scalability: Ensuring low-latency, million-transaction throughput required microservices and event-driven architecture.
  • Engagement Fraud Detection: Building robust fraud-risk assessments to detect bot-driven engagement.
  • Regulatory Compliance: Navigating RBI guidelines for UPI, GST/TDS for influencer payouts, and ASCI disclosure rules.

Accomplishments that we're proud of

  • First-mover engagement-first influencer tiering in India with rupee-based pricing.
  • Real-time instant cashback payouts using UPI’s low-cost rails.
  • 4-layer RBAC with performance-based dynamic permission escalation.
  • PWA, enabling offline browsing and native-like UX without an app store.
  • Deploying a hybrid database strategy for transactional consistency and high-volume analytics.

What we learned

  • True influencer value lies in quality engagement and conversion, not just follower count.
  • Event-driven designs drastically improve throughput and fault isolation for high-volume fintech flows.
  • Building regional adaptability—languages, cuisines, festival seasons—drives user adoption.
  • PWA with ISR balances SEO and performance for content-heavy platforms.

What’s next for DineConnect

  • Launch AI-powered influencer-restaurant matching marketplace with real-time bidding
  • Expand to 10+ additional Indian cities, add regional languages
  • Introduce blockchain-backed cashback audit trails for transparency
  • Add crypto payouts (e.g., USDC) for premium influencers
  • White-label solutions and API marketplace for third-party integrations

Built With

  • Languages & Frameworks: Next.js, React, TypeScript, Go, Gin, gRPC
  • Cloud & DevOps: Docker, Kubernetes, AWS/GCP, Istio, Terraform, GitHub Actions
  • Databases & Caching: PostgreSQL, Redis, MongoDB, ClickHouse
  • APIs & Messaging: Instagram Basic Display API, Razorpay/Cashfree, Stripe, RabbitMQ
  • Analytics & ML: Prometheus, Grafana, Python ML microservice for attribution

Try it out

Live Demo

Code block example

puts "Welcome to DineConnect!"


Learn more about our technical deep dive in the repo’s Wiki and API documentation.

Built With

Share this project:

Updates