Zyft AI
An AI-powered system for delivering personalized emotional touchpoints (greetings today), currently in a production phase with limited rollout, while stabilizing email delivery and scaling reliability.
Product overview
Problem
- Employee birthdays, festivals, and milestones are often missed due to manual tracking.
- Creating personalized messages + visuals is time-consuming and requires effort.
- Generic mass emails feel transactional and reduce emotional connection.
What Zyft solves
- Centralizes employee data (name, email, DOB, department, interests) to enable personalization.
- Generates AI-written messages and themed greeting visuals on demand.
- Enables an end-to-end workflow: employee data → personalized card → email delivery.
Current MVP: AI Greetings (Emotion Engine v1)
- Add employees (supports bulk import via CSV/XLSX).
- Occasion engine highlights today’s birthdays + upcoming events.
- Select an occasion (e.g., Christmas, Diwali, New Year, Eid).
- AI generates a themed card image + personalized message (regen supported).
- Send to individuals or batch send via email delivery pipeline.
System architecture
Current deliverable (MVP)
- User touchpoint: Admin Dashboard (current MVP interface).
- Trigger: Admin initiates greeting generation for an occasion.
- Emotion Engine v1: AI Greeting Cards module orchestrates the flow.
- Shared intelligence layer: pulls personalization signals from the Unified User Profile (likes/interests).
- Gemini API (LLM): generates greeting message + image content for delivery.
Future expansion packs
- Mobile app (future): Voice command interface as an additional touchpoint.
- Voice intent decoder: NLP to interpret gift intent and context.
- Marketplace aggregator: Amazon/Shopify APIs to fetch products.
- Unified User Profile: remains the central personalization source (cards now, gifts later).
- Gemini API: extends from greetings to recommendations and wish generation.
2025 Progress Overview
Production-focused snapshot of delivery, outcomes, and what’s next.
Stage
Production phase (limited rollout)
Demos / Reviews
10–15 internal reviews (est.)
Teams / Users
2 teams • ~20–30 users (est.)
Emails / Runs
~1,000–1,500 sends (est.)
Time saved
~3–5 min per greeting • ~1–2 hrs per campaign (est.)
Stability
~93–97% successful sends (est.) — improving
Production readiness
Deployment
Done
Secrets/config
Done
Logging
Done
Retries/idempotency
In progress
Deliverability
In progress
Alerting/metrics
Next
Email delivery stabilization
Current focus
- Resolve delivery edge cases impacting consistency.
- Reduce failures via stronger retry + provider error handling.
- Improve visibility into send outcomes and failure reasons.
Next actions
- Idempotency + retry with backoff for batch sends.
- Delivery status instrumentation (success/fail reason codes).
- Deliverability checklist (domain alignment + warm-up if needed).
Milestones
Project kickoff & scope alignment
Formal kickoff, scope finalization, and alignment on MVP goals, success criteria, and delivery plan.
MVP foundation delivered
Established core project structure, admin UI skeleton, and baseline backend workflows.
Core workflows implemented
Completed employee directory and occasion-based greeting flow with AI generation integration.
Email delivery & batch sending
Integrated email delivery pipeline with batch send support; began stabilization and reliability improvements.
Production phase (limited rollout)
Entered production phase with internal rollout; focused on resolving remaining email delivery bugs and hardening the system.
What shipped
Frontend
Backend
AI
Infra
2026 Roadmap
Next 30 days
- Stabilize email delivery (retries, idempotency, edge-case handling)
- Add delivery status tracking + failure reason reporting
- Finalize rollout checklist + runbook
Next 60 days
- Improve personalization quality (context inputs + templates)
- Expand template library for festivals + milestones
- Introduce admin controls (limits, audit trail, basic governance)
Next 90 days
- Scale internal rollout to additional orgs/teams
- Add scheduling/automation for reminders and campaigns
- Operationalize monitoring (alerts, dashboards, SLO targets)
Impact / traction
Current stage
Production phase (limited internal rollout)
Internal adoption
2 teams • ~20–30 internal users
Replace with exact count when available
Email sends (to date)
~1,000–1,500 test + pilot sends
Includes batch runs and retries
Generation volume
~300–600 greetings generated
Message + card image generation
Operational impact
~1–2 hrs saved per campaign (est.)
Manual copy/design + coordination reduction
Known gap
Email delivery bugs (stability + deliverability) in progress
Key outcomes & learnings
- Personalization matters: strong signal that tailored messages outperform generic communication.
- AI reduces effort: automated copy + visuals significantly reduce manual work.
- Data unlocks potential: centralized employee data enables multiple future engagement use cases.
- Clear expansion path: from greetings → gifting → shared experiences.
Revenue model
Core: Freemium → Premium
- Free tier: basic templates + limited monthly sends
- Premium tier: automation, advanced templates, analytics, governance
- Target pricing: $15–$25/month (per admin/seat)
Expansion: B2B / Enterprise
- Target partners: companies running employee engagement programs
- Value: higher retention + lower CAC via org-wide rollout
- Target pricing: $2K–$5K/org/month
$10K/month math (reference model)
300 Premium users × $20 ARPU = $6,000/month + 2 Enterprise clients × $2,000/month = $4,000/month → $10,000/month.
Team
Harshini Raja
Backend Developer
Supports backend feature development and integrations; ensures reliability, validation, and production-ready service behavior.
Pavan Kumar Srinivasulu
Backend Developer
Supports backend API endpoints and data flows; contributes to system stability, performance, and integration readiness.
Rohit Parthiban
Backend Developer
Contributes to backend services and deployment-readiness; supports integrations, debugging, and operational improvements for smooth delivery.
Dhanasekar Ravi Jayanthi
Frontend Developer
Owns core UI development; builds and refines user-facing flows to ensure a clean, usable one-page experience for stakeholders.
Pujitha Kallu
Frontend Developer
Develops frontend components and page structure; focuses on clarity, content organization, and maintainable UI patterns.
2026 business plan & roadmap
- Expand from festival greetings to full AI-powered individual greetings using shared moments.
- Utilize shared moments to build robust emotion-based messages to improve personalization.
- Integrate marketplaces (Amazon, Shopify, Etsy) for end-to-end fulfillment.
- Focus on scalable B2B adoption and recurring enterprise usage.
Tech stack
Zyft product stack (current MVP)
- Backend: Python + Flask (REST APIs)
- Frontend: React
- AI: Google Gemini (text + image generation)
- Email: Resend (single + batch)
- Database: PostGresSQL + JSON logs
- Deployment: GCP (Ubuntu)