OVERVIEW
PROCESS
DESIGN
IMPACT
OVERVIEW
PROCESS
DESIGN
IMPACT

Funda

Designed for high-engagement, short-form learning

Funda

Designed for high-engagement, short-form learning

For

Funda

Role

Lead Product Designer

Timeframe

May 2025 - April 2026

Tool Used

Chat GPT, Magnific

Funda

Designed for high-engagement, short-form learning

For

Funda

Role

Lead Product Designer

Timeframe

May 2025 - April 2026

Tool Used

Chat GPT, Magnific

PROJECT

Overview

Overview

Overview

Funda is an edutainment app launched in June 2025 with a goal to reach ₹1 Cr revenue by Dec 2025. Content velocity was the primary growth lever.

Funda is an edutainment app launched in June 2025 with a goal to reach ₹1 Cr revenue by Dec 2025. Content velocity was the primary growth lever.

Funda is an edutainment app launched in June 2025 with a goal to reach ₹1 Cr revenue by Dec 2025. Content velocity was the primary growth lever.

Approach

Approach

Built an AI-assisted thumbnail generation system that:

  • Automates composition

  • Maintains visual quality

  • Scales output exponentially


We don’t need better designers.
We need a system that designs

Problem Statement

Problem Statement

Video production scaled from 100 → 1000 videos/month

  • Only few editors + 1 designer

  • Each video required:

    • Thumbnail (high impact)

  • Banner (high volume)

Thumbnails drive clicks. But we couldn’t design fast enough to grow.


Video production scaled from 100 → 1000 videos/month

  • Only few editors + 1 designer

  • Each video required:

    • Thumbnail (high impact)

  • Banner (high volume)

Thumbnails drive clicks. But we couldn’t design fast enough to grow.


Challenges

Challenges

Thumbnails directly influence CTR → growth → revenue
Manual design couldn’t scale at required speed.
How might we create high-quality, high-performing thumbnails at massive scale without increasing team size?

Thumbnails directly influence CTR → growth → revenue
Manual design couldn’t scale at required speed.
How might we create high-quality, high-performing thumbnails at massive scale without increasing team size?

DESIGN

PROCESS

Process

Process

followed

followed

Design in Phases

1. Early Exploration (Phase 1 — GPT)

  • Used GPT’s image model (April 2025)

  • Solved key limitation: accurate text rendering

  • Created structured prompts using:

    • Composition logic

    • Wireframes

    • References

Breakthrough:

  • Iteratively trained a GPT chat for stable outputs

  • Generated a reusable master prompt system

Impact:

  • ~30 thumbnails/day





2. Scaling Output (Phase 2 — Ideogram)

  • Identified GPT limitations in layout/design quality

  • Shifted to Ideogram for:

    • Better typography

    • Stronger compositions

Workflow:
GPT → structured prompt → Ideogram → final output

Impact:

  • ~2x increase in production speed

  • Improved visual quality




3. Brand Layer (Phase 3 — Real Influencers)

  • Transitioned from generic characters → real influencer faces

Challenge:

  • AI models distorted faces → broke brand trust

Solution:

  • Integrated Google’s Nano Banana for:

    • High-fidelity face generation

    • Consistency across thumbnails

Outcome:

  • Stronger recognition

  • Higher engagement potential




4. Quality Upgrade (Phase 4 — December Push)

  • Revamped older thumbnails + videos

  • Standardized quality across library

Result:

  • ~2000 high-quality videos

  • Achieved ~90% of ₹1 Cr target (~₹90L by Dec 31, 2025)




    Video Thumbnails (typography Based)




5. Systemization (Phase 5 — Automation)

Built a repeatable pipeline:

  • Freepik Spaces

    • Pre-trained influencer character models

    • Structured inputs → consistent outputs



  • Bulk generation (ChatGPT + App Script)

    • Used for simple banners

    • Optimized for cost + speed






Design in Phases

1. Early Exploration (Phase 1 — GPT)

  • Used GPT’s image model (April 2025)

  • Solved key limitation: accurate text rendering

  • Created structured prompts using:

    • Composition logic

    • Wireframes

    • References

Breakthrough:

  • Iteratively trained a GPT chat for stable outputs

  • Generated a reusable master prompt system

Impact:

  • ~30 thumbnails/day





2. Scaling Output (Phase 2 — Ideogram)

  • Identified GPT limitations in layout/design quality

  • Shifted to Ideogram for:

    • Better typography

    • Stronger compositions

Workflow:
GPT → structured prompt → Ideogram → final output

Impact:

  • ~2x increase in production speed

  • Improved visual quality




3. Brand Layer (Phase 3 — Real Influencers)

  • Transitioned from generic characters → real influencer faces

Challenge:

  • AI models distorted faces → broke brand trust

Solution:

  • Integrated Google’s Nano Banana for:

    • High-fidelity face generation

    • Consistency across thumbnails

Outcome:

  • Stronger recognition

  • Higher engagement potential




4. Quality Upgrade (Phase 4 — December Push)

  • Revamped older thumbnails + videos

  • Standardized quality across library

Result:

  • ~2000 high-quality videos

  • Achieved ~90% of ₹1 Cr target (~₹90L by Dec 31, 2025)




    Video Thumbnails (typography Based)




5. Systemization (Phase 5 — Automation)

Built a repeatable pipeline:

  • Freepik Spaces

    • Pre-trained influencer character models

    • Structured inputs → consistent outputs



  • Bulk generation (ChatGPT + App Script)

    • Used for simple banners

    • Optimized for cost + speed






FINAL

Designs

Hi- Fidelity

Hi- Fidelity

Design Mocks

Design Mocks

A scalable pipeline:

Input → Prompt → Model → Output

  • Content idea

  • GPT structured prompt

  • Model selection:

    • Ideogram → design-heavy thumbnails

    • Nano Banana → real faces

    • GPT → bulk/simple banners

  • Output: ready-to-use creatives


IMPACT &

Outcome

Impact

Impact

Created

Created

Scale

  • 100 → 1000 videos/month

  • ~10x growth without increasing team size

Production

  • Manual → semi-automated system

  • Hundreds of creatives generated weekly

Quality

  • Maintained high CTR-focused design

  • Improved consistency across content

Business Impact

  • ~2000 videos published

  • ₹90L revenue by Dec 2025 (~90% of target)

  • Scaled to ₹5.6 Cr monthly revenue by March 2026

Content + thumbnail system became a key growth engine for revenue acceleration

Overview

Process

Design

Outcome

OVERVIEW
PROCESS
DESIGN
IMPACT
OVERVIEW
PROCESS
DESIGN
IMPACT

HOME

HOME

WORK

WORK

CONTACT

CONTACT

LinkdIn

LinkdIn

Behance

Behance

Instagram

Instagram

Overview

Process

Design

Impact

Create a free website with Framer, the website builder loved by startups, designers and agencies.