Skip to content
Surinder T.

Surinder ThakurAI Product Designer and Builder between Dubai and India, for UAE, US and EU teams

I build AI products people actually use.

Custom tools and agents for your operations, built so your team still uses them a month later.

Book a scope call

1 slot open for Q4 2026

Working prototype in 2 weeks · $3,500

Built and led over 19 years

B2B agenticDutch B2B commerce. Blueprint prototype.Live example (opens in a new tab)AI SaaSProUX. RAG audit platform, built solo.350+ UX principles (opens in a new tab)Mobile appPrahar. Agents plan your day.Agentic workflow (opens in a new tab)D2C commercePuffy. Design manager, $1B brand.Measured+67% YoY revenue

+21%mobile checkout conversion· Measured, A/B test

“Designs implemented have had positive impact on user experience and bottom line growth.”
Arthur Andreasyan · Founder, Puffy
Surinder at work on a laptop, on a bench at Dubai Marina with the Ain Dubai wheel behind him

Trusted since 2007.

Fintech, commerce, edtech, healthcare, real estate, SaaS and mobile apps. Now I bring that to AI tools and agents teams keep using.

  • Puffy
  • Whisper Invest
  • Alucor
  • Enov8
  • Medellin
  • BidOnHomes
  • Ramsha Homes
  • AIBA World
  • Olam Solutions
  • Zoom Errands
  • LivOn4ever
  • ooliv

What I build.

The 4 kinds of product I build most, each with an agent at work.

  • Fintech

    New customers quit halfway through signup.

    An agent reads each document and fills the form, so customers type less. You check only the fields it flags.

  • Commerce

    Buyers still order by email and phone.

    A buying agent turns each message into an order, so buyers keep their routine. It reaches your ERP only after you approve.

  • Operations

    Your team still copies data between tools.

    Agents run invoices, approvals and onboarding across your tools, by your rules. Anything over your limit waits for you.

  • Mobile apps

    People install the app, then stop opening it.

    An AI layer plans each user's day, so the app earns its place. It says what it moved, and why.

Built, shipped, in use.

Each one links to the real thing, so you can judge the work, not the pitch.

Dutch B2B commerce · Blueprint prototype, 2 weeks

Agents take the order. People keep the decisions.

An urgent order in minutes, not an hour, Modeled, from the prototype

Try the prototype (opens in a new tab)

One urgent order, from call to ERP.

  1. The buyer calls and says the order
  2. The agent asks 1 question, never a form
  3. It checks live stock and the contract price
  4. Over €300, the workshop owner approves
  5. Written to the ERP, with 10 minutes to undo

From the prototype, with mock ERP data.

An urgent order in minutes, not an hour.

  • The buyer orders by voice, WhatsApp or email
  • Sales reps keep the hard calls and the relationship
  • Brakes, steering and airbags always go to a person

Each agent has 1 job. People approve rules and exceptions.

An agent orders alone only when all 3 checks pass.

The buyer’s 8 steps. People stay in control at 3.

The merchant’s side: every gap found before the first order.

“Your agentic prototype was strong: the buyer and merchant flows were coherent… your human-in-the-loop approach to trust stood out.”
Product teamDutch B2B commerce platform

AI SaaS · Built solo, live at proux.design

Agents do the research. Every answer cites its source.

Every answer cites its source, across 350+ UX principles

Try ProUX (opens in a new tab)

Not a chatbot. You set a task and ProUX runs it.

Your product gets the same engine. Your rules, your data, decisions it can defend.

The ProUX engine reads 5 sources:

  • Experts with curated skills.
  • Your project context.
  • Guideline library.
  • Your screen.
  • The ProUX method.

It returns the right expert, the right model and cited design decisions.

You set the task. It picks the expert and the model.

  • Every choice named on screen
  • One click changes it

The task: Audit this product page. Buyers drop before Add to cart. What do I fix first?

  • Expert: Design audit. You asked for an audit.
  • Model: Auto · Sonnet 5. Deep, picked by the job.
  • Grounding: 3 project documents. Your project first.

Then it runs: Analysing the design, Searching ProUX guidelines, Mapping findings.

It used 16 guideline sources.

Every claim links to its source.

  • 1 tap opens the guideline and its evidence
  • When no source covers it, it says so

Darken the price text: it measures 2.43:1 against 4.5:1.

Source 1: guideline 411, text legibility in eCommerce.

Collapse long specs, so Add to cart stays in view.

Source 2: guideline 32, collapse sections.

Assumption: no project document says who buys.

It measures the screen. It does not guess.

ProUX measures a product page from its pixels.

  • Price text: 2.43:1, needs 4.5:1.
  • Add to cart: 36pt, needs 44pt.
  • Body text: 16px, passes.

The score is 64, at risk: 100 less 24, 8 and 4.

Price text is fixed at 5.20:1. Add to cart is fixed at 48pt.

With 2 criticals fixed, the score moves from 64 to 80.

Same task. A different kind of result.

    The same task, answered by a general LLM and by ProUX.

    • LLM: “Looks light”. ProUX: 2.43:1 against 4.5:1, Measured from pixels.
    • LLM: “Best practice”. ProUX: Guideline #411, 1 tap away.
    • LLM: 8 equal tips. ProUX: Ranked fixes, Effort on each.
    • LLM: A new answer. ProUX: Same rules, Every page.
    • LLM: Start over. ProUX: Re-measured, Fixed. Now 5.20:1.

    Puffy · design manager, 2 years

    Repositioned as attainable luxury.

    +67% YoY revenue, Measured year on year, validated by A/B test

    Now it looks worth the price.

    • Emotion: dark, warm, muted gold reads as luxury.
    • Social proof: 4 named awards, then 5 stars from 12,000+ sleepers.
    • Zero price: $240 of gifts, free.
    Puffy's product page before: a cream page, a long paragraph, a summer sale box, a grid of sizes, and 2 buttons.
    Puffy's product page after: the Cloud mattress in a dark, warm room, 4 named awards and 5 stars from 12,000+ sleepers above the fold, $240 of gifts free with every mattress, and 1 add to cart with a money-back guarantee under it.

    +67% YoY revenue. Measured year on year, validated by A/B test

    Checkout cut from 3 steps to 2.

    Mobile checkout, step 1 of 2.

    • Anchoring: $3,438 struck through.
    • Trust: the total stays pinned, so nothing surprises at pay.
    • Zero price: each gift shows its full price, then free.
    Step 1 of Puffy's 2-step mobile checkout, in a shopper's hand: $1,788 to pay with $3,438 struck through, the order summary pinned at the top, the Lux mattress at $899, and free luxury bedding gifts, each at its full price, then free.

    Every model, side by side.

    • FullStory and user tests: shoppers left after the hero, unable to compare models.
    • Decision fatigue: the facts line up.
    • Effort: 1 size picker for all 4.
    Puffy's compare page: the Cloud, Lux, Royal and Monarch mattresses side by side under 1 size picker. Lit on each card, what changes between them: the feel, from medium-firm to luxury-plush; the height and layers, from 10 inches and 6 layers to 16 inches and 12; and the price, each with its list price struck through.

    Ads that look like the store.

    • Message match: the ad's dark, warm look is the store's.
    • Benefit first: each ad leads with what the sleeper gets.
    • 2 pillars: materials and comfort.

    The first fold, tested every month.

    Find the friction, test the fix, keep the winner.

    “Designs implemented have had positive impact on user experience and bottom line growth.”
    Arthur AndreasyanFounder, Puffy

    Pursely · money management app, design concept

    Spend, budget and invest. All in 1 app.

    40+ screens, light and dark, 1 design system, A design concept, sold as a template

    Spending and investing, 1 toggle apart.

    • Effort: Manage and Invest share 1 toggle at the top.
    • Feedback: income, expense and shift sit under the curve.
    • Anchoring: returns show in green beside the value.

    The month's spend, before it runs out.

    • Goal gradient: a gauge fills as the month's spend grows.
    • Mental accounting: each budget shows what is left of its limit.
    • Defaults: a new budget repeats monthly from today.

    Stocks in a list, the detail 1 tap down.

    • Feedback: each stock shows its move in green or red.
    • Effort: 6 ranges sit under 1 chart, from 1 day to all.
    • Progressive disclosure: the full chart waits behind 1 button.

    Identity first, then the money.

    • Trust: you pick where the code goes: email or mobile.
    • Effort: face or fingerprint, set up once at sign-in.

    40+ screens, light and dark.

    • Consistency: 1 set of styles drives both themes.
    • Rhythm: an 8 pt grid sets every gap and size.
    “If he says he will do something, there's no question that it will get done.”
    Avin KlineCEO, Lucyd

    Purpose · learning app concept, 30+ screens

    Built for the phone. Nobody learns alone.

    From first run to live class, in light and dark

    Try the live demo (opens in a new tab)

    The right course from the first screen.

    • Effort: search, categories and picks for you on the first screen.
    • Goal gradient: your courses show as a path, each with its progress.
    • Habit: a Today screen puts the next lesson first.

    Video, files and talk in 1 lesson.

    • Focus: the video on top, then chapters, files and discussion in 3 tabs.
    • Goal gradient: done chapters tick off, and a bar fills under the one playing.
    • Effort: each chapter shows its length and downloads for later.

    Questions answered beside the lesson.

    • Belonging: questions and answers sit under the lesson, not in another app.
    • Social proof: likes and replies show which answers helped.

    The first run asks what you want to learn.

    • Clarity: 1 picture and 1 line say what the app is for.
    • Relevance: you pick your subjects first, so home starts with them.
    • Effort: interests are bubbles you tap, no typing.

    Know who teaches before you start.

    • Authority: followers, reviews and students sit under the name.
    • Liking: a short About me, in the teacher's own words.
    • Choice: all 34 of their courses, 1 tap away on See All.

    3 reasons people quit learning apps, each answered.

    • Friction: every screen drawn at 375 wide, never a shrunk desktop page.
    • Goal gradient: progress on every path and chapter, so the end feels close.
    • Belonging: questions and replies sit inside the lesson, under the video.

    30+ screens, 1 design system.

    • Consistency: every screen is built from the same parts, so it reads as 1 app.
    • Comfort: light and dark, from the same system.
    • Rhythm: every space sits on the 8pt grid.
    “I have been working with Surinder for almost a decade… very professional and always on-point with meeting deadlines and expectations.”
    Carlos AriasFounder, Medellin.co

    MyHouz · design manager at RGG, US real estate

    Map-first search. 1% back on every home.

    MVP live in 90 days, 150 buyer inquiries in month 1

    Map first, with the cashback on every home.

    • Effort: the map and the homes side by side, filters above both.
    • Anchoring: 1% cashback next to 0% at Redfin and Zillow.
    • Salience: every home card shows its cashback in green.

    The price, then what comes back.

    • Mental accounting: the price also shows per month.
    • Gain framing: the cashback sits in green under the price.
    • Effort: 2 next steps: see it in person, or make an offer.

    The monthly cost, before any call.

    • Mental accounting: $4,099 a month, split into 5 parts.
    • Defaults: 20% down and a 30-year loan, filled in.
    • Trust: taxes, insurance and HOA fees shown, none left out.

    The same homes, plus 1% back.

    • Anchoring: $8,500 back, beside $0 at Redfin, Zillow and an agent.
    • Concreteness: the 1% also said in dollars, an average of $8,500.

    Every filter in 1 panel.

    • Effort: every filter in 1 panel, home types as icons.
    • Feedback: a price histogram shows where the homes are.
    • Gain framing: the price range shows the cashback it earns.

    3 methods found where buyers slowed down.

    • Competitor scan: Zillow and Redfin lead with the map, sticky filters and a visible perk.
    • Search-flow audit: a list-first, 3-step path reset filters and added taps on mobile.
    • Buyer interviews: 6 of 8 first-time buyers would try MyHouz for the 1% alone.
    “Effectively led the implementation of layout variations, showing good judgment in selecting and prioritizing test concepts that align with our business objectives.”
    HimanshuDirector of Marketing, RGG

    Findout · travel listing app, design concept

    Find a place. Know it before you go.

    32 screens for hotels, restaurants, events and places, A design concept, sold as a template

    Try the live demo (opens in a new tab)

    What you want, then where.

    • Effort: 4 kinds of place as icons, before any typing.
    • Defaults: the location has a locate button, so nearby is 1 tap.
    • Focus: 1 navy button, Search, so the next step is plain.

    Every place on the map, with its reviews.

    • Effort: the list and the map are 1 tap apart, filters beside them.
    • Social proof: every card leads with its stars and its number of reviews.
    • Recognition: wifi, parking and no smoking show as icons.

    Each kind of place gets its own next step.

    • Anchoring: the price range shows before any date is picked.
    • Certainty: an Opened badge on the photo, before the trip.
    • Effort: 1 main action for each kind, like Reserve table, Get quote or Buy ticket.

    The score first, then a way to add yours.

    • Social proof: 5.0 from 84,559 ratings, above every review.
    • Reciprocity: Post review sits under the score, so readers give back.
    • Effort: a search inside the reviews finds the ones that matter.

    Dates and guests, set in taps.

    • Recognition: dates and guests stay on top of every booking step.
    • Effort: rooms, adults and children as plus and minus, no typing.
    • Defaults: 1 room and 2 adults to start, children at 0.
    “He consistently delivers on time often ahead of schedule and his professionalism is unmatched.”
    Carlos AriasFounder, Medellin.co

    Signal · news app concept, 30+ screens

    The day at a glance. The full story in 1 tap.

    30+ screens and 80+ components, day and night, Design concept

    Try the live demo (opens in a new tab)

    The day at a glance, from 1 screen.

    • Scannability: topic tabs on top, then 1 lead story, big.
    • Context: today's date and weather sit beside the news.
    • Social proof: every story shows how many readers liked it.

    A story page built for the long read.

    • Focus: 1 photo, the title and the author, then the text.
    • Rhythm: a pull quote breaks up the long read.
    • Effort: follow and share sit right under the title.

    Live coverage, update by update.

    • Urgency: a red LIVE tag sits above the headline.
    • Recency: each update carries its date and time.
    • Effort: share sits under every update.

    The feed starts with what you pick.

    • Defaults: topics come first, so the first feed already fits.
    • Clarity: each topic says in 1 line what it covers.
    • Control: Skip is always there.

    30+ screens, in light and dark.

    • Comfort: a night theme for reading in the dark, from the same parts.
    • Reach: right to left in 1 switch, for Arabic and Urdu readers.
    “If he says he will do something, there's no question that it will get done.”
    Avin KlineCEO, Lucyd

    Used and trusted.

    Most AI builds work in the demo, then people go back to spreadsheets.Mine are built for the weeks after launch.

    Most AI builds against my builds, row by row
    Most AI buildsMy builds

    Money: Act on their own

    Pay inside your rules, with a 10 minute undo

    Answers: Give no source

    Show the source, so your team can check

    When unsure: Guess

    Stop and ask a person

    Success: Measured at launch

    Measured by weekly use

    Book a scope call

    Free 30 minute call, no pitch. Then a 2 week Blueprint: a working prototype you can click.

    People never say where a product hurt them. They just leave.

    So I own every project myself, from first sketch to live product. No account managers and no handoffs.

    Surinder at work on his laptop in a café in Dubai
    Who designs it
    I do. 19 years in UX, HFI and NN/g certified.
    Who builds it
    I lead and build it myself. When the work needs more hands, I bring in senior specialists I trust.
    Who you talk to
    Me, every week, with a live demo of what changed.
    What I build with
    • Claude
    • Claude Code
    • Cursor
    • Figma
    • Next.js
    • Vercel
    • Supabase
    • Voyage AI
    • Cohere Rerank

    Most of my clients found me through a referral.

    More about me

    In their words, on camera.

    • “Their quality is phenomenal. They deliver what they say they're going to do.”
      Avin KlineCEO, Lucyd, USA
    • “He's very collaborative and easy to work with. He wants to produce what's best for you and not what's best for him.”
      Sheila RaperMD, AJElite Homes, USA
    • “Thank you very much for delivering awesome artwork over the last 10 years.”
      Uli SchönleberFounder, Ooliv, Germany
    • “There's just never anything that is a bother to them. They are listening all the time.”
      José NavarroFounder, Barefoot Media, Australia
    • Designer I trained
      “Before this, I would hand off a Figma file and hope for the best. Now I own the whole process.”
      Phulvinder SinghDesign lead, Ikokas Technologies
    • Designer I trained
      ProUX
      “I grew more in this program than I did in the last two years, learning on my own.”
      Priyanka GillSenior UI/UX designer
    • Designer I trained
      ProUX
      “My manager noticed within the first week. Two months later I got promoted.”
      Ritul ChatterjeeSenior UI/UX designer, Grazitti Interactive
    • Designer I trained
      ProUX
      “I thought I was keeping up, but this program showed me the gap between using AI casually and professionally.”
      Ashish KhuranaSenior UI/UX designer

    Fixed prices, in order.

    Each price is fixed in writing before work starts.

    1. Blueprint

      Start here

      2 weeks

      Founding price, $3,500usual price, $5,000

      Credited to your build

      • A business case, in numbers
      • A working prototype you can click
      • A build proposal with a fixed price and timeline

    2. Build

      Fixed in your proposal

      From $20,000

      Typically $30,000 to $60,000

      • Led by 1 senior, start to finish
      • A live demo every week
      • Scope and price fixed in writing

    3. Retainer

      Monthly

      Founding price, $8,000usual price, $10,000

      Month to month

      • Launch with you
      • Track weekly use after launch
      • Keep improving what people use

    Founding rate for the first 5 clients.

    A lower price, for a case study after launch.

    • $1,500 off the Blueprint
    • $2,000 off each retainer month, for 12 months
    • Case study shown only after launch
    • Anonymous, if you ask
    Book a scope call

    Free 30 minute call, no pitch. Scope and price within 24 hours.
    Prices in USD. No hourly billing.

    See full pricing

    Before you ask.

    Straight answers on risk, privacy and price.

    Something else on your mind?

    Ask Surinder

    • Generic chatbots, design only projects and hourly work. I take on custom AI tools and agents with a clear business goal. A tool with no goal has nothing to measure, so nobody can say it worked.

    • Agents pay only inside the rules you set. Each payment waits 10 minutes before it goes out, so you can undo it. Anything outside the rules waits for a person.

      Read full answer

    • Yes. Nothing is shown before launch, so no competitor sees it early. Founding clients agree to a case study after launch. Your name and any sensitive data can stay anonymous.

    • I am taking 5 clients at a lower price. In return, each agrees to a case study after launch. The Blueprint is $3,500 instead of $5,000. The retainer is $8,000 a month for your first 12 months, then $10,000. The offer ends at 5 clients, so no countdown rushes you.

    • From $20,000. Most builds land between $30,000 and $60,000. The Blueprint ends with a fixed price in writing. You know the full cost before you commit.

      Read full answer

    • The Blueprint is paid upfront, and all of it is credited to your build. The build is paid in 2 or 3 milestones, set in your proposal. You pay each one when it is done and you have seen it working. There is no separate deposit, because the Blueprint is already credited. The retainer is billed monthly and runs month to month.

    • The Blueprint takes 2 weeks. Build time depends on scope, so the proposal fixes the timeline before you commit. You see a live demo every week until launch.

      Read full answer

    • It covers work on your live tool after launch. Each week we agree what ships, and you see it live. I track weekly use, fix what slows people down and add small features. A new tool or a big feature gets its own fixed proposal. You see its price before any work starts. It runs month to month, with 30 days' notice. You should stay for results, not a contract.

    • That is a good result. You keep the business case and the prototype. Learning it cost $3,500. A build would have cost $30,000.

      Read full answer

    • Yes. Most of my client work has been for US teams, found through referrals. You get a live demo every week, whatever your time zone.

    Prefer to talk it through?

    Not yet? See a real Blueprint first. Proof, no pressure.

    The case study and the working prototype, made for a Dutch B2B commerce platform.

    See the Blueprint example (opens in a new tab)Or book a scope call

    On time every time

    “If he says he will do something, there's no question that it will get done.”
    Avin Kline · CEO, Lucyd