Hi, I am Lohith!

A Product Designer seasoned in
end-to-end web design — from spec to shipped.

Portrait

More from the Desk!

A look at the range — not just the highlights.

Product Design

End-to-end web apps — from first user journey to production.

View curated work →

Web & AI Build

Landing pages, marketing sites, and AI-assisted builds — volume at quality.

View curated work →

Instructional & Branding

Learning systems, visual identity, and content that teaches.

View curated work →

DesignForge

A 6-phase spec-to-ship AI-human methodology. Delivery time: months → weeks.

Most AI-assisted builds start fast and end slow — the speed gains get eaten by rework. DesignForge is how I prevent that. A 6-phase spec-to-ship framework: gap analysis upfront, three distinct design directions per component, and a deviation rule that makes every meaningful drift from spec a formal decision. AI generates. I curate. Nothing gets delegated silently.

Delivery time

1–3 months

↓

1–2 weeks

The 6-phase process

0

Journey Planning

Map the feature — user journeys, component inventory, and what to prioritise first.

→
1

Sandbox Creation

Set up the isolated build environment with mock data before writing a single line of production code.

→
2

System Rough Build

Build the complete rough system without production constraints. Front-load the gap discovery.

→
3

Design Iteration

Explore three distinct design directions freely, separate from production, before committing to one.

→
4

Sandbox Integration

Implement the chosen design with the full design system and token constraints applied.

→
5

Prod Migration

Audit frontend and backend, generate a Token Deviation Report, migrate to production.

3-option

Exploration rule

Every component of any significance goes through three genuinely distinct design directions before one is chosen. Not three variations of the same direction — three different approaches. The choice is the curation. This is where designer judgment does its job.

1
✓
3

±20%

Token deviation rule

When AI output diverges more than 20% from the spec's expected scope, a formal Token Deviation Report is triggered. The designer either accepts the deviation with a stated reason, or course-corrects. No change gets absorbed silently. Every meaningful departure from intent is a documented decision.

token
±20%

4 production apps. 6 months. Same designer, spec to shipped.

Neuron

3× quiz completions, 10× question attempts

PACE

Complete rebuild done in days instead of months

SPARK

One day of manual build → feed a JSON

Website v3

Ranks on previously unranked keywords

Precision and Creativity

Skills that I bring to table

Product & UX/UI Design
Figma
Adobe XD
Web Design & Development
Webflow
Framer
HTML/CSS
React
AI Stack
Claude Code
Claude.ai
Figma AI
Lovable
Replit
Marketing & Branding
Photoshop
Illustrator
Canva
Premiere Pro
Camtasia
Instructional Design
Storyline 360
Rise 360
Technical
Python + Selenium
Prompt Engineering
SQL
API Integration
*Logos are properties of respective companies
Expert Learning

What they say?

Echoes of Impact!

Design is collaborative. From mentors to teammates, words that remind me — "I've made an impact!"

S
"Working with Lohith on the public website redesign changed what I thought was possible from a single person. He didn't just design — he owned the content, the structure, and the production, end to end. We shipped faster and cleaner than I've seen on any project of that scale. When I look at the current site, I see his decisions in every section."

Sundeep Eddu

Head of Marketing and Sales, e-GMAT