Building BizEcho – A WhatsApp-First Customer Feedback Platform
An architectural breakdown of a serverless feedback collection and reputation management platform built with AWS, SQS, Node.js, and WhatsApp Cloud API.
Project
BizEcho
Role
Full Stack Developer
Duration
6+ Months
Tech Stack
Node.js, TypeScript, AWS Lambda, API Gateway, DynamoDB, SQS, WhatsApp Cloud API, React, Serverless Framework
About BizEcho
Businesses often struggle to collect actionable customer feedback. Traditional email surveys suffer from low response rates, while negative customer experiences frequently go unnoticed until they impact retention and online reputation.
BizEcho was built to solve this challenge by enabling businesses to collect customer feedback directly on WhatsApp, automate engagement workflows, monitor customer sentiment, and improve online reviews through a centralized platform.
The Challenge
Before BizEcho, partner businesses faced severe hurdles in tracking client happiness:
- [!] Low response rates from email-based feedback systems.
- [!] Lack of real-time visibility into customer satisfaction.
- [!] Manual follow-ups for dissatisfied customers, causing operational delays.
- [!] Difficulty converting positive customer experiences into public online reviews.
- [!] No centralized system for managing customer feedback across multiple branches and locations.
The Solution
I helped build a highly scalable, serverless feedback management platform that automates customer engagement through the WhatsApp Cloud API.
Key Features
⚡ WhatsApp Feedback Automation
Triggers automated feedback requests immediately after a service is completed. Customers rate their experience interactively using WhatsApp buttons, generating personalized conversation paths based on their scores.
📊 Customer Sentiment Tracking
Aggregates Net Promoter Score (NPS), Customer Satisfaction (CSAT), and text reviews into a centralized management dashboard to plot satisfaction trends over time.
🛡️ Smart Review Management
Intelligently redirects highly positive responders to public platforms like Google Reviews to build social proof. Conversely, negative comments are routed internally to support agents for instant resolution.
📍 Multi-Location & Branch Support
Enables branch-wise performance metrics, permitting corporate managers to benchmark individual store performance, view local ratings, and export reports.
🔄 Lead & Engagement Workflows
Maintains follow-up queues for non-responders, structures message intervals safely, and tracks complete user engagement analytics.
Technical System Design
To handle bursty campaign volumes and maintain near-zero server infrastructure costs, we designed an event-driven serverless architecture on AWS.
Cloud Backend & Database
The core backend is implemented using Node.js and TypeScript on AWS Lambda, exposed via AWS API Gateway REST endpoints. SQS (Simple Queue Service) buffers incoming webhook events from WhatsApp, processing messages asynchronously to prevent API rate limits. Amazon DynamoDB serves as a fast, horizontally-scalable key-value store to manage active conversation states, business records, and customer surveys.
WhatsApp Integration Engine
Integrates with Meta's WhatsApp Cloud API. The service validates and matches payload responses (interactive buttons, quick replies) against active survey flows. Includes automated validation for approved message templates to prevent spam blocks.
Analytics & Frontend Dashboard
A single-page React application providing corporate managers with real-time feedback feeds, branch-level reporting, custom campaign configurators, and customer communication widgets.
Engineering Challenges & Solutions
Scaling Webhook Message Processing
When feedback campaigns were triggered, thousands of customers received notifications simultaneously. The resulting webhook events from WhatsApp (delivered, read, button clicks) flooded our API Gateway, causing backend spikes and threat of data loss.
Solution: Decoupled ingestion from processing. Webhooks immediately push raw events to an AWS SQS queue and respond with a 200 OK. Downstream consumer Lambda functions process queues at a controlled concurrency limit, utilizing Dead Letter Queues (DLQ) for failed messages with automated retries.
WhatsApp Template & Payload Restrictions
WhatsApp enforces strict parameters on message structures, requiring pre-approved templates and strict formatting. Slight payload deviations result in silent message drop-offs or API errors.
Solution: Created a template configuration engine that automatically maps business parameters to Meta-compatible JSON payloads, runs schema validation locally before sending requests, and logs delivery failures via callback status tracking.
DynamoDB Analytical Queries Degradation
As the volume of customer responses crossed hundreds of thousands, reading and aggregating CSAT/NPS ratings location-wise became sluggish, leading to database scans and dashboard delays.
Solution: Restructured DynamoDB partition and sort keys using Global Secondary Indexes (GSIs). Implemented aggregate database write-sharding, and pre-calculated hourly summaries using a DynamoDB Stream trigger, keeping dashboard response times below 150ms.
My Contributions
As a key developer on the project, my responsibilities covered the complete engineering lifecycle:
- > Designed, developed, and deployed serverless REST APIs on AWS Lambda.
- > Built custom integrations with the WhatsApp Cloud API using webhooks.
- > Designed DynamoDB data models and configured GSI keys for analytical aggregates.
- > Configured SQS messaging queues and DLQ error fallback rules.
- > Built components in the React dashboard, implementing dashboard analytics visualizations.
- > Setup automated deployment pipelines using the Serverless Framework.
- > Collaborated with Product Owners and QA teams to formulate end-to-end integration tests.
Project Impact
📈 Business Impact
- • Tripled customer feedback response rates compared to email.
- • Automated 90% of follow-ups, reducing manual overhead.
- • Accelerated public review acquisition by 40% in key branches.
- • Allowed branch managers to resolve complaints before negative reviews hit Google.
⚙️ Technical Impact
- • Zero base-server maintenance overhead due to Serverless structure.
- • Scaled seamlessly to handle campaigns sending 50,000+ texts.
- • Cut database query times from 2s down to under 150ms.
- • Robust webhook queue processing with near-zero message loss.
Key Learnings
Developing this platform provided highly valuable technical and architectural insights:
- [01] How to design robust, backpressured, event-driven integrations with third-party webhooks.
- [02] Structuring schema models for DynamoDB to optimize for real-time dashboard aggregation queries.
- [03] Implementing strict validation for high-throughput messaging channels to comply with platform policies.
- [04] Balancing user interaction speed, batch job scheduling, and operational cost budgets.
WhatsApp & Serverless Consulting in Pune
Collaborating locally in Pune and building high-scale custom web services for Indian enterprises.
From my development office in Pune, I deliver modern communications platforms that align with the high standards expected by local startup ecosystems in Hinjawadi, Baner, and Kalyani Nagar. Being located in Pune, I offer direct collaboration, workshops, and architectural review sessions to build robust software.
If your business needs a senior software developer to integrate customer review workflows, build custom web app dashboards, or move servers to efficient AWS Lambda architectures, I offer high-caliber full-stack engineering tailored for immediate business impact.
Want to Build Something Similar?
Whether you need a specialized WhatsApp API integration, serverless APIs on AWS, or a database-tuned React dashboard, let's connect and discuss your requirements.