I design the trust layer of AI products.
The surface where a person decides to act on the model — or override it. Fifteen years of measured outcomes, with the baselines to prove them.
Data-literate at the model layer. I shape what gets measured. I ship the front-end. · Founding · Staff IC · Director-level · Available · 4 weeks' notice
Arpit Maheshwari
Design leader · AI & data-intensive products
Based in Indore, India · fully remote, with 4–5 hours of daily overlap with US East Coast.
Selected Work
Fifteen years leaves more NDA than portfolio. What's here is representative. What isn't — four cases — is on request.
PTC University — Learning Connector
Multi-platform learning consolidation across 11 languages. The redesign wasn't the hard part — the political case for killing four products was.
Telefónica MyO2 & Priority Moments
Two O2 UK consumer products — the MyO2 self-service app (4M+ users) and Priority Moments loyalty (2.6M sign-ups). Designed and built every screen, mobile web.
Four more · NDA · Walk-through on request
Programmatic Advertising Platform
Confidence-first design for media buyers who wouldn't act on the algorithm's calls.
AI-Assisted Private Equity Investing
Refused to ship the model without an "explain this score" surface. PE analysts started using it.
OrgOS · Transparent Org Tooling
Eight modules letting a 200-person team self-organise. Zero managers; transparency as coordination.
Technical Due Diligence Platform
AI-extracted technical signals + partner-grade trust. Cycles compressed from 3 weeks to 4 days.
Patterns
AI Design Patterns Library
Patterns I've shipped into production for AI products across multiple industries. Each one carries the tradeoff I learned the hard way — what I tried, what broke, what I changed.
The Act / Review / Ignore rule
One score, one action. A confidence number that doesn't tell the user what to do next is an opinion in percentage points. Every confidence surface I ship maps to exactly one of three actions — act, review, or ignore — never a bare 87%.
Read the rule →Confidence Score Patterns
5 core patterns for displaying AI confidence: numeric, color-coded, language-based, and composite approaches.
AI Failure States
How to communicate uncertainty, errors, and limits transparently when AI systems can't deliver.
ML Explainability
Patterns for making ML decision-making visible to non-technical users without overwhelming them.
Human-in-Loop Patterns
Interface patterns for keeping humans in control when AI handles high-stakes recommendations.
How I Lead
I'm the design lead who reads your model evals, sits in on your customer calls, and opens PRs for the components I own. By the end of week one, I've written the design diagnosis your team didn't know they needed.
What you're hiring me to own
- The trust layer
- The product surface where users decide whether to act on the model — or override it.
- The design language
- The component system and patterns the next designer inherits — written down, not in my head.
- The ML/UX contract
- The explicit agreement between the ML team and the user about what the system can and can't do.
How I partner with each function
- With engineering
- I pair on eval design before I touch the UI. I open PRs for the CSS I own. When eng and I disagree about feasibility, I default to the cheaper experiment first.
- With product / founder
- I push back on the roadmap when the data says we're wrong. I commit to outcomes, not deliverables. I write the design doc; you write the spec.
- With customers
- I sit in on five calls in week one and one a week after. I read the support tickets myself. I won't ship an AI feature until I've watched a user fail with it.
When I've been wrong
I argued the recommendation card should show three ranked options. Engineering pushed for one — the top-ranked, no alternatives. I lost the argument. In the next quarter's A/B test, the one-option version converted 2.3× better — three options made users freeze. I changed how I design every recommendation surface after.
What I won't do
- I won't be the only designer past 40 people. Past that, design becomes an org problem and I want a teammate, not martyrdom.
- I won't ship an AI feature without a designed failure state. Non-negotiable. If your team disagrees, we're not a fit.
- I won't run design ops in parallel with shipping product. Pick one. Doing both badly is the most common senior-designer trap.
What the people I've worked with say
Over the past four years at Talon, Arpit has been instrumental in shaping four distinct products from the ground up. His user-focused designs are remarkably intuitive yet adept at handling complex workflows… If you need a designer who excels at combining strategic vision with practical execution, Arpit is the person to call.
Arpit has worked with me for years and I value his honesty and hard work. He's been an integral part of my staff… involved in all facets of the team, from design to development to hiring and onboarding of new members.
Arpit teams up with designers very well, not only does he flawlessly execute the UI implementations but he pushes back on design decisions using his UX expertise… I'd recommend Arpit to any team looking to improve their final product.
About
Fifteen years across EdTech, telecom, AdTech, fintech, and org tooling — the last six fully remote across three companies, with millions of users shipped. Based in Indore, India. I write The Trust Layer, a newsletter on designing AI products people actually act on.
On my first accessibility project I spent a week with my monitor switched off, navigating by screen-reader — designing for someone who isn’t you starts by becoming them.
Writing
The Agentic MVP: Why Your Next Launch Will Be Lovable, Not Just Viable
Every founder knows the pit in their stomach on Launch Day. How the rise of agentic systems is rewriting what "minimum viable" means — and why lovability is now the bar.
The AI Fight Club: Weaponizing Claude and Gemini for Bulletproof Products
Pitting AI systems against each other to strengthen product robustness. A practical method for stress-testing your AI features before users do it for you.
The New Renaissance: How AI is Transforming Us from Software Operators to Digital Artisans
Ushering in a new era of digital entrepreneurship. How AI is changing what it means to build, and what designers must understand about the tools reshaping the industry.
Contact
Hire me to lead design.
Founding · Staff IC · Director-level · Available · 4 weeks' notice · Fully remote (GMT+5:30) · 4–5 hr daily overlap with US East Coast
Book a 30-min intro callNot hiring full-time? I keep a few hours a month for advisory and short trust-layer audits — ask.
Building an AI product and want fresh eyes first? Send a link on LinkedIn — within 48 hours I'll reply with a short Loom: three concrete improvements, free, no pitch.
What it's like working with me
Based in Indore, India (GMT+5:30). 4–5 hours daily overlap with US East Coast. Six years, three companies, millions of users shipped — without being in the same room as anyone on the team.