Skip to main content
< Sai Komal/>

Sai Komal

Product Designer & UX Engineer

Open to roles


  • Research
  • Design
  • Prototype
  • Ship
Years
3+

Years

Manual coordination cut
40%

Manual coordination cut

Case studies
3

Case studies

Scroll
Experience
3+ years
Focus
Enterprise SaaS & AI products
Industries
Fashion tech · Logistics · DevOps · Government
Front-end
React, Next.js, Tailwind
Design
Figma, design systems
AI workflow
Claude & AI-assisted workflows
Location
India · open to remote
Availability
Open to new roles
Resume
FIG. 01Product thinking

Three disciplines, one system.

Not three handoffs on the way to a product. One loop, and the experience is what comes out of it.

UX

Logistics, DevOps, fashion-tech

  • Research
  • User needs
  • Flows
  • Interaction

Engineering

B.Tech Mechanical, then the front end

  • Systems
  • Constraints
  • Code
  • Feasibility

AI

Claude & AI-assisted workflows

  • Context
  • Intelligence
  • Decision
  • Confidence

Product experience

One system, not three handoffs

FIG. 02Construction

An interface is assembled, not drawn.

An interface is not drawn, it is assembled. This is the order the pieces go in.

  1. 01IdeaBrief · Canvas · Grid
  2. 02StructureSidebar boundary · Header boundary · Workspace · Navigation
  3. 03DesignHeader controls · Summary cards · Figures · Filters · Table
  4. 04SystemStatus · Utilisation · Distribution · Activity
  5. 05ProductSelection · Detail panel · Interaction · Product

Demonstration interface — information architecture from the Kudra case study. Values shown are UI samples, not project results.

FIG. 03AI product

Intelligence is a loop, not a feature.

Intent, context, a decision, an action, an outcome — and the feedback that changes what happens the next time.

  1. 01

    User intent

    What someone is actually trying to get done

    • Stated request
    • Implied goal
    • Constraints
  2. 02

    Context

    What the product already knows before it thinks

    • User data
    • Task context
    • System state
  3. 03

    AI system

    Interprets, weighs the options, commits to one

    • Reasoning
    • Decision
    • Generation
    • DataRecords, history, state
    • ToolsRetrieval, scoring
    • InterfaceWhere it can act

    Low confidenceAsk, don't guess

  4. 04

    Action

    What the product does with the decision

    • Interface
    • Workflow
    • Recommendation

    GuardrailValidated before it is shown

  5. 05

    Outcome

    A result the user can interrogate, not just accept

    • Result
    • Confidence
    • Undo
  6. 06

    Feedback

    Accepted, corrected or abandoned — and the system keeps it

    • Accepted
    • Corrected
    • Refined

Returns to context

Conceptual stages — a product architecture, not a model one. The two branches are where the design work actually is: ask when confidence is low, and validate before anything reaches a screen.
FIG. 04Problem solving

Nothing arrives in order.

Nine inputs, none of them sequenced. Getting them into something buildable is most of the job.

Nine inputs, no order

  • User
  • Problem
  • Research
  • Business
  • Tech
  • Constraints
  • Data
  • AI
  • Workflow

Sorted into three

  • EvidenceWhat is actually true
    • User
    • Problem
    • Research
  • BoundaryWhat is not negotiable
    • Business
    • Tech
    • Constraints
  • MaterialWhat there is to build with
    • Data
    • AI
    • Workflow

Which becomes a structure

  1. 01

    Insight

    Interviews · Empathy map

  2. 02

    Model

    Problem statement · IA

  3. 03

    Flow

    Flows · Edge states

  4. 04

    Design

    Design system · Components

  5. 05

    Product

    Working prototype · Usability findings

The nine inputs are conceptual. The five stages are not — they are the five steps of the method this practice actually runs on, named for what comes out of each, and the artefact under every one is that step's own recorded output.
FIG. 05The design machine

A problem goes in. A product comes out.

The same five steps, drawn as the apparatus they actually make up — a problem in one end, a shipped product out of the other.

  1. Input01

    Discover

    Talk to the people who will actually use it.

    Interviews · Empathy map

  2. Filter02

    Define

    Scope it — what is in, what is out, what success means.

    Problem statement · IA

  3. Form03

    Design

    Wireframes to hi-fi, on a system rather than one-off screens.

    Flows · Design system

  4. Construct04

    Build

    Wire the real thing up in React, not a static file.

    Working prototype

  5. Test05

    Validate

    Real usage decides what gets fixed next.

    Usability findings

  6. Output

    Product

    In front of the people it was built for, doing the job.

    Kudra · fleet operations

    Kudra fleet management screen — the live vehicle view from the case study
The machine is the five steps of the method — Discover, Define, Design, Build, Validate — drawn as apparatus. Every description and artefact is that step's own; the screen at the delivery end is Kudra's fleet view from the case study.
03Track record

Data, then design, then AI products.

Four roles, one direction of travel — data, to design, to owning the product.

  1. UX Designer

    Onliest

    May 2026 — Present

    Visakhapatnam, India

    • Designed REX, an AI recommendation product — requirements through flows, interaction models and prototypes.
    • Designed the AI edge states: low-confidence, no-result, cold-start, loading, empty and error.
    • Built reusable Figma patterns across Marketplace, HR Admin and Revenue using Variables and Variants.
  2. Product Owner & UX Designer

    ONDL

    Oct 2024 — May 2026

    Visakhapatnam, India

    • Cut trip-creation effort 20% and manual operational effort 30% by consolidating the workflow.
    • Designed Kudra end to end — trip creation, driver assignment, vehicle allocation, fuel tracking.
    • Ran field research with drivers, dispatchers and fleet supervisors.
  3. UX Designer

    Sweya Infotech

    May 2023 — Oct 2024

    Visakhapatnam, India

    • Designed enterprise products from research and IA through to high-fidelity UI.
    • Ran interviews, usability testing and stakeholder workshops to find workflow breakage.
    • Created reusable components and design-system patterns for scalable interfaces.
  4. SQL Developer

    Oracle

    Nov 2021 — Apr 2023

    Hyderabad, India

    • Built enterprise data-integration and reporting solutions across APIs and business workflows.
    • Shipped internal reporting tools that widened access to business data.
04Contact

Let's talk.

Looking for a full-time Product Designer or UX Engineer role. The form is the fastest reply.