# Parul Chouhan — Product Designer

> Portfolio site: https://parulchouhan.com
> Contact: parul.chouhan003@gmail.com
> LinkedIn: https://linkedin.com/in/parulchouhan
> Location: Bengaluru, India

## About Parul Chouhan

Parul Chouhan is a product designer with 4+ years of experience building complex B2B SaaS platforms from 0 to 1. She specialises in AI products, design systems, and enterprise UX. She was previously the founding product designer at Intelekt AI.

She thinks in systems, not just screens. She started as a Graphic Designer and UI Designer, but soon discovered her love for UX. Her work sits at the intersection of design, product strategy, and execution, shaped by experience both as a Product Designer and a Chief of Staff.

She has delivered over 10+ products across industries including B2B SaaS, AI-driven products, e-commerce, Travel-tech, and B2C products. She is a daily user of AI tools including Claude, Figma AI, and Cursor.

### What drives her

She believes great products aren't just about aesthetics and visuals — it's about usability, creating seamless experiences, and measurable outcomes. She asks questions before sketching screens, listens before designing flows, and always pushes to question assumptions rather than reinforce them.

She read Don Norman's "Design of Everyday Things", which helped shape her journey and her love for solving complex design problems.

---

## Work Experience

### Intelekt AI — Founding Product Designer
**Duration**: 2024 – 2025

Joined as the sole product designer at an enterprise voice-first AI startup. Built the entire design system from scratch across 7 product modules, designed onboarding flows, and shaped the product strategy from early stage through growth.

Key achievements:
- Built a token-first design system across 7 product modules that reduced design-to-development handoff time by 40%
- Designed an outcome-led onboarding system that improved early activation by 30 percentage points (42% → 72% of users completing first AI agent call within 72 hours)
- Reduced manual demo dependency for the founding team by enabling self-serve product discovery

### Chief of Staff (Previous role)
Operated across business strategy and execution alongside design work, giving her a unique perspective at the intersection of product design and business outcomes.

### Freelance Product Designer
Delivered multiple consumer and B2B product designs for clients across India and internationally, including CULUX (luxury fashion e-commerce), travel platforms, and fintech products.

---

## Case Studies

### Case Study 1: Improving early activation by 30% with outcome-led onboarding design

**URL**: https://parulchouhan.com/case-study/improving-early-activation-by-30-with-outcome-led-onboarding-design
**Client**: Intelekt AI
**Tags**: B2B | Product Design | Strategic Thinking | AI-Driven Design
**Role**: Product Designer
**Duration**: May – June 2025
**Team**: Me (Product Designer), 2 developers

#### Company Context

Intelekt AI is an enterprise-grade voice-first AI platform that automates customer conversations using intelligent agents. It powers autonomous voice interactions that can qualify leads, resolve queries, update CRMs, and create tickets without manual intervention. The product targets operations and CX teams inside BFSI, Hospitality, Retail, and Utilities companies.

By 2024, the platform had a working product and a growing sales pipeline. But every new client required a manual demo from the founding team before they could experience the product. There was no self-serve path.

#### Problem

The team was manually running demos for every new enterprise client. Users who signed up independently had no guidance. They landed on a blank dashboard with no clear first action. Without experiencing the AI agent directly, users lacked the confidence to configure anything or invite their team. The product could not scale without consuming the founders' time on every single new account.

The core tension: Onboarding a voice-first AI platform is not like onboarding a form builder. Users are asked to trust a system they cannot fully see before it makes a real phone call on their behalf. Too much structure → users hesitate. Too little → teams struggle with clarity and trust.

User needs identified:
- Start instantly
- Experience the "aha moment" quickly
- Later grow into teams and shared workspaces

#### Discovery

Signal came from observing founders run demos, sitting in on onboarding calls with early clients, and reviewing drop-off patterns from the first cohort of self-serve signups via session recordings.

Key findings:
1. Users hesitated when asked to invite teammates before they had made a single call themselves → Value must be demonstrated before collaboration is introduced
2. Users could not understand what the AI agent was capable of from the dashboard alone — they needed to hear it to believe it → The aha moment was auditory; the first experience had to be a live call, not a product tour
3. Early signups attempted to configure real agents immediately, triggering unintended downstream actions → The first interaction needed to be sandboxed — real enough to impress, controlled enough to be safe

#### Goal

Move users to the product's aha moment as quickly as possible by letting them experience an AI agent in action, rather than explaining how it works. Replace the manual demo with a self-serve experience that makes users feel confident rather than overwhelmed.

Primary metric: Percentage of new users who completed a first successful AI agent call within 72 hours of signup.

#### Design Strategy: 4 Key Decisions

**Decision 01**: Users experience a live AI agent interaction before being asked to invite teammates, configure access, or navigate further.
- Why: Most users don't understand the product well enough early on to make structural decisions. Confidence in the product should precede commitment to teams or access.
- Principle: Action before explanation — users learn faster by experiencing outcomes than by configuring systems.
- Outcome: Faster arrival at aha moment, fewer abandoned onboarding flows, team expansion happened after clarity.

**Decision 02**: Every new user starts in a personal workspace by default.
- Why: Early exploration is individual, not collaborative. Shared setups too early introduce anxiety.
- Principle: Personal context before shared context — ensuring ownership is clear before collaboration expands.
- Outcome: Immediate product usage, clear ownership from first interaction, reduced setup hesitation.

**Decision 03**: The first AI agent interaction was intentionally scoped, constrained, and executed in a controlled environment.
- Why: Voice interactions have real-world cost and consequences. Early users lack sufficient mental models. Misconfiguration is expensive and trust-breaking.
- Principle: Confidence before enforcement.
- Outcome: Zero unintended calls during onboarding, reduced operational risk, increased confidence to proceed with deeper configuration.

**Decision 04**: Access ownership was implicitly established during onboarding by making the workspace creator the Admin, without asking users to configure roles or permissions explicitly.
- Why: During onboarding, users lack context to make correct access decisions. Explicit role configuration early increases cognitive load.
- Principle: Clarity over flexibility — it's better to be clearly right than flexibly wrong.
- Outcome: Reduced permission-related confusion, fewer support queries, access complexity introduced only after users were ready.

#### Business Impact

- Demo dependency dropped significantly — founding team no longer required on every new client call
- Inbound lead quality improved — prospects who booked sales calls had often already completed a test call
- Support load on onboarding decreased significantly

#### Results

| Signal | Before | After |
|---|---|---|
| First agent call completed within 72 hours | 42% of signups | 72% of self-serve signups (+30 percentage points) |
| Teammates invited after first call | Rare | Consistent pattern across first cohort |
| Onboarding-related support requests | High | Meaningfully reduced within 4 weeks of launch |

#### Conclusion

Onboarding is not an entry flow — it is a product strategy decision. The 30% activation improvement was not the result of a better UI. It was the result of understanding that a voice-first AI product earns trust through demonstration rather than explanation, and designing an entire system around that single insight.

---

### Case Study 2: Built a design system that reduced design to development time by 40%

**URL**: https://parulchouhan.com/case-study/built-a-design-system-that-reduced-design-to-development-time-by-40
**Client**: Intelekt AI
**Tags**: B2B | Design System | Token Architecture | AI-Native Patterns
**Role**: Sole Product Designer
**Duration**: 3 months
**Team**: 1 Design + Engineering team
**Tools**: Figma, FigJam, Notion

#### Company Context

Intelekt AI is a voice-first agentic AI platform that automates enterprise contact center operations. The product lets operations teams configure AI voice agents that can qualify leads, resolve customer queries, update CRMs, and trigger downstream workflows without human intervention.

Joined as the founding product designer. Within the first two months it became clear that the pace of feature development was creating a problem that would compound badly if left unaddressed. The design system was not a visual project — it was a risk mitigation decision.

#### Problem

By the time the audit began, 7 product modules were being designed and built in parallel. Each module had evolved independently under pressure to ship.

Issues found:
- The Campaign module and Agent Configuration module both contained data tables built separately with different row heights, sort indicators, empty states, and hover behaviors
- Form inputs had 3 different border radius values across the product
- Buttons had 4 different padding specifications
- Error states were handled inconsistently: some showed inline validation, some toast notifications, some neither
- Every time engineering built a new feature, they made micro-decisions about spacing, color, and component behavior from scratch

#### My Role

Owned this project entirely alongside regular product design workload:
- Auditing all modules to catalogue inconsistencies and define token architecture
- Building the component library in Figma from atoms through organisms
- Writing documentation for every component, including states and edge cases
- Working directly with engineers during implementation to close interpretation gaps

#### System Architecture

Built using Atomic Design principles, adapted for a complex enterprise product:

| Layer | Responsibility | Examples |
|---|---|---|
| Atoms | Tokens, primitives, foundational styles | Color, spacing, typography |
| Molecules | Functional UI blocks | Forms, filters, inputs |
| Organisms | Reusable product patterns | Tables, modals, navigation |
| Templates | Layout and structure | Page scaffolding, grid system |
| Pages | Real product states and edge cases | Empty states, error screens |

#### Token Foundation

Built token-first system to align design and engineering. Included:
- Color tokens (semantic naming, dark/light mode support)
- Typography scale (standardised across all 7 modules)
- Spacing scale (8px grid system)

#### What Was Built First and Why

Prioritised by leverage — which components, if standardised, would have the highest impact on the most surfaces immediately:

1. **Buttons, inputs, and form fields** — appeared in every single module. Standardising these first eliminated the most frequent source of inconsistency across the entire product in a single pass.
2. **Data tables** — highest inconsistency, used across Campaign Management, Leads, and Analytics modules.
3. **AI-native patterns** — components invented from first principles: Live Call Transcript Component (streaming text, speaker attribution, confidence indicators, timestamp markers) and Workflow Trigger Outcome States (timeline pattern with semantic status tokens).

#### Before and After

| | Before the system | After the system |
|---|---|---|
| Campaign data table | Custom row height, bespoke sort icon, no empty state | Standardised table organism, shared sort indicator, documented empty state with action |
| Form inputs | 3 different border-radius values | Single token, consistent across all 7 modules |
| Error states | Inline, toast, or nothing depending on module | Standardised: inline validation on all form inputs |
| Button padding | 4 different specs across the product | One spec per variant, documented and tokenised |

#### Results

| Signal | Before | After |
|---|---|---|
| Slack clarifications per handoff | 8 to 15 messages over 2 to 3 days | 1 to 3 messages, resolved same day |
| Topics of clarification | Spacing values, hex codes, states, edge cases | Genuine product ambiguities requiring design judgement |
| QA inconsistency flags | Frequent across modules | Substantially reduced within 4 weeks of adoption |

The 40% reduction in handoff friction was an observed and consistent signal across three months of product development after system adoption.

#### Conclusion

A design system built under startup conditions is not a pristine Figma file. It is a series of decisions made under pressure, with incomplete information, while simultaneously shipping the product it is meant to serve. The value was in the agreement it created. Tokens gave design and engineering a shared language. Component documentation replaced repeated conversations with a single written source of truth.

---

### Case Study 3: Designing a premium omnichannel fashion marketplace for India's HNIs

**URL**: https://parulchouhan.com/case-study/designing-a-premium-omnichannel-fashion-marketplace-for-indias-hnis
**Client**: CULUX
**Tags**: Freelance | Consumer Product | Luxury E-commerce
**Role**: Freelance Product Designer
**Duration**: 6 to 8 weeks
**Team**: Solo + Founders

#### Company Context

CULUX is India's premium omnichannel fashion platform, bridging the gap between India's rich craft heritage and modern luxury consumers. The project involved designing an MVP for a premium e-commerce experience that celebrates traditional craftsmanship while meeting the sophisticated expectations of contemporary premium shoppers.

The founding team brought deep industry expertise. This wasn't a speculative startup — it was an experienced team addressing a real market need.

#### The Problem

India's 10M+ HNIs are growing fast but every existing platform forces them to choose between curation, experience, and speed. Less than 5% of platforms offer any personalisation in premium commerce.

The brief: take the product from zero to a live MVP in 6 to 8 weeks. No existing design system, no brand guidelines, no design team.

#### Market Research

- 100M+ affluent consumers by 2027 in India
- Less than 5% of platforms offer personalisation in premium commerce
- 19%+ CAGR in accessible luxury in India
- 3x YoY premiumisation growth
- Gen Z & Millennials demand experience-led shopping, not transactional commerce

#### Design Research

The client facilitated access to 25 HNI users across 5 cities (Delhi, Mumbai, Bangalore, Chennai, Hyderabad). Competitive analysis across COYU, Ciceroni, Myntra, and Nykaa.

Key observations translated to design decisions:

| What I observed | What I designed for |
|---|---|
| HNI users spent more time navigating than shopping on existing platforms | Lead with curation at every entry point. The platform should feel chosen, not surfaced. |
| Time scarcity was a stronger frustration than price or selection gaps | Every flow learnable in under 2 minutes. Friction is a trust failure, not a usability problem. |
| Most browsing happened on mobile between meetings and during travel | Mobile-first without compromise. Touch as capable as desktop, not a reduced version. |
| Heritage and craftsmanship were described as trust signals, not decorative content | Brand story before product grid. Story earns the transaction. |
| Too many options with too little context was the primary abandonment driver | Restraint is a product strategy. Show less, mean more. |

Four design principles derived from synthesis:
1. **Curation over completeness** — Fewer products with more context. The user should feel a stylist selected this, not an algorithm surfaced it.
2. **Time is the luxury** — Every interaction faster and clearer than expected. Friction signals the platform does not respect their time.
3. **Story before transaction** — Brand story before add-to-bag. Trust is built through narrative before purchase intent is requested.
4. **Mobile-first without compromise** — Touch interactions as capable as desktop. Not a stripped-down version of the full experience.

#### Information Architecture

Three key IA decisions:
1. **Curations leads, not Categories** — Putting Curations first in the navigation signals editorial intent immediately.
2. **Occasion-first category taxonomy** — Categories use Party, Festival, Vacation, Wedding, Workwear as primary filters rather than product types like Dresses and Tops. HNI users shop for occasions, not product types.
3. **Brand pages as destinations** — A brand page on Culux is not a filtered result set. It is an editorial destination with the designer's story, craftsmanship context, and curated collection.

#### User Flow Decisions

1. **Quick view before commitment** — Research showed HNI users evaluate 4 to 6 products before committing to a detail page. Quick view gives full product detail from the listing page without navigation, reducing back-navigation abandonment.
2. **Occasion filter before brand filter** — A user who has selected Wedding before seeing a single product is a user who already knows what they are looking for.
3. **Styling consultation as a flow entry point, not an exit** — Personal styling is surfaced on the homepage and at the top of category pages, making subscription value visible before commitment.

#### What Was Deliberately Not Built in V1

| Excluded | Reasoning |
|---|---|
| Loyalty points and tier display | HNI users are not motivated by points mechanics. Loyalty through curation is more consistent with the positioning. |
| Social sharing from product pages | This user group values privacy in purchases. Social sharing felt inconsistent with the premium, discreet positioning. |
| Full AI personalisation engine | Requires behavioural data that does not exist at MVP. Hand-curation is more honest at this stage. |
| Approval workflows for brand access | Would introduce friction at the point of highest intent. |

#### Core Insight

In luxury, restraint is a product strategy. Every decision to show less, to curate rather than display, to surface one brand story rather than fifty product thumbnails, came from research. India's affluent crave identity and personalisation in their choices. They do not want more options. They want better ones, chosen for them, with the story of why those choices matter.

---

## Other Work

### Tourzee – Your Travel Partner
**URL**: https://parulchouhan.notion.site/Tourzee-Your-Travel-Partner-deceef9391964e80ae06134a95672628
**Type**: Personal project | Travel and Hospitality | 2023
A travel companion app designed to simplify trip planning for solo and group travellers.

### Slack Onboarding Teardown
**URL**: https://parulchouhan.notion.site/How-Slack-s-onboarding-turns-first-time-users-into-active-teams-2f2bee283b758078a59ad385275d8202
**Type**: UX Teardown | Onboarding | SaaS | 2023
An in-depth analysis of how Slack's onboarding turns first-time users into active teams.

### Designing a notification system for an all-in-one finance app (Fino)
**Type**: Product Design | Fintech | Systems Thinking | Personal — Coming Soon

---

## Skills & Tools

- **Design**: Figma, FigJam, Prototyping, Design Systems, Component Libraries
- **Research**: User Interviews, Usability Testing, Competitive Analysis, Session Recording Analysis
- **AI Tools**: Claude, Figma AI, Cursor, ChatGPT
- **Methods**: Jobs-to-be-Done, Atomic Design, Heuristic Evaluation, Information Architecture
- **Domains**: B2B SaaS, Enterprise UX, AI Products, E-commerce, Fintech, Travel Tech

---

## Site Pages

- **Homepage**: https://parulchouhan.com/
- **About**: https://parulchouhan.com/about
- **Work / All Projects**: https://parulchouhan.com/work
- **Shots (UI Gallery)**: https://parulchouhan.com/shots
- **Case Study — Onboarding**: https://parulchouhan.com/case-study/improving-early-activation-by-30-with-outcome-led-onboarding-design
- **Case Study — Design System**: https://parulchouhan.com/case-study/built-a-design-system-that-reduced-design-to-development-time-by-40
- **Case Study — CULUX**: https://parulchouhan.com/case-study/designing-a-premium-omnichannel-fashion-marketplace-for-indias-hnis

---

## Contact

- **Email**: iparulchouhan@gmail.com
- **LinkedIn**: https://linkedin.com/in/parulchouhan
- **Resume**: https://parulchouhan.com/parul-chouhan-productdesigner.pdf
- **Location**: Bengaluru, India
- **Open to**: Senior Product Designer and Staff Product Designer roles