Accessibility and AI Product Designer
I designed AI tools for disability employment at Our Ability, then built them in Python and TypeScript.
Before Our Ability I was the designer and engineer on a QR menu product for restaurants. MS Information Systems, Syracuse University.
- Based
- Chennai, India
- Focus
- Accessibility and AI products
- Builds in
- Python, TypeScript, React
- Open to
- Full-time, freelance, relocation
Selected work, (6)
Redesigning a disability job platform that was hard to use if you had a disability
Cutting a four-feature redesign down to two, because our users can't absorb change the way most people can.
- Desktop screens
- 57
- Tablet screens
- 59
- Color pairs pass contrast
- 22 of 22
OrangeWork Hub: replacing paper shift scheduling for Syracuse student workers
A mobile app for student shift scheduling, built solo after I lived the paper process myself
- Friends swapped a shift on the first try
- 8/8
- To schedule a shift. About 30 minutes on paper
- Under 30 sec
- Screens designed
- 38
Knowt: Designing a streak system that makes students come back every day
Design a streak and rewards system that gives students a daily reason to return, without leaning on guilt the way Duolingo does.
- Screens designed
- 36
- Flows
- 9
- Layout options explored
- 13
PlatePost: Designing a video-first menu platform for restaurants
One template, six restaurants, and one usability fix that changed all of them
- Restaurants on one template
- 6
- Months as the only designer
- 21
- Customers scanned in a 4 week test
- 1 in 5
Abli.ai: Making PDFs readable for screen readers
An AI pipeline that rebuilds inaccessible PDFs, benefit guides, HR forms, legal documents, so screen readers can read them properly.
- Failed checks on a sample flyer
- 2 → 0
- Issues compared across two AI audits
- 34
Self-Disclosure and Accommodation
A tool that helps candidates search jobs, ask questions about disclosing a disability, and get accommodation guidance.
- Status
- Prototype
- Platform
- Web
- Role
- Technical Product Designer
Kind words, (3)
All recommendations on LinkedIn (opens in a new tab)Manoj performed a wonderful task in bringing our vision into reality. What I saw in my head with our system, Manoj was able to create the template for our coders to build. Manoj performed exemplary work!
John RobinsonCEO, Our Ability, Inc.Worked with Manoj on the same team, January 2023 Manoj is one of the best team members I’ve met. His dedication to learning new things, being actively engaged in conversations, and having a growth mindset tells me he’ll do well wherever he ends up.
At PlatePost, Manoj has directly been in charge of creating dozens of VideoMenus that have collectively garnered 100,000 customer views.
Brandon ShiFounder, PlatePostManaged Manoj directly, April 2025 He demonstrated exceptional skills in translating the client's vision into an intuitive and visually appealing website.
I highly recommend him for UX/UI design opportunities, as his commitment and reliability set him apart and make him an invaluable asset to any team.
Anjuman AlamAssociate Project Manager, iConsult CollaborativeManaged Manoj directly, July 2023
Things I've built, (4)
Shipped
This site's AI assistant
Ask about my work and it answers from my case studies, resume and LinkedIn, then moves the page to the section that proves it. It says when it doesn't know instead of guessing, and it works with just a keyboard.
Built with: Next.js · Claude Haiku 5.5 · Anthropic API
Debrief
Paste a meeting transcript and Claude pulls out the action items, decisions and open questions. You check everything first. Then it creates Notion tasks, drafts a Gmail recap, posts to Slack and books a follow-up. Two days later it wakes up on its own and nudges anyone whose task is still open.
Built with: Next.js · Vercel Workflow · AI SDK · Claude
Walk in Their Shoes
Upload a screenshot of your design and see it the way people with different kinds of vision do. An AI model then reviews it and tells you what to fix. I built it in a day.
Built with: Replit · AI review
Building now
Motor Archive
A used car tool I'm building now. The idea: you enter a VIN, it pulls the car's recalls, complaints and specs from NHTSA data, and Claude turns that into a plain-English verdict on whether it's worth buying. After you buy, it helps you make sense of a check engine light.
Built with: React · Node.js · Express · MongoDB · NHTSA API
