I design AI products that make complex work simple.

Deep in product thinking, grounded in user intuition.

Selected case studies

Work that moved real numbers.

A free report that became a $300K+ product.ARR, Preflight and SmartScreen
  • Best Author Tool, SSP 2026

Paperpal / Preflight and SmartScreen

Authors were failing journal rules they never knew existed. Preflight turned that hidden rulebook into a plain verdict before submission, free to the author, with paid AI edited fixes.

B2CAIPublisher AdminMonetization
Read the case study
Screening workflows behind a ~$600K ARR contract.2M+ manuscripts supported

Springer Nature and Taylor & Francis / Manuscript Integrity Screening Platform

Led end to end UX design and scaling for an AI manuscript integrity platform. The workflows contributed to a 3 year Springer Nature contract (~$600K ARR).

Protected by NDA

Get in touch to learn more about it.

AI review for 100K manuscripts a year.Helped win $150K a year contracts

American Medical Association / AI Manuscript Integrity and Peer Review

Designed AI solutions supporting 100K manuscripts a year. Contributed to contracts totaling $150K a year for 2 years, plus $100K for 1 year.

Protected by NDA

Get in touch to learn more about it.

A simpler checkout that lifted conversion from 17% to 24%.Researcher.life multi product checkout

Researcher.life / Checkout Redesign

Redesigned the multi product checkout journey by simplifying information hierarchy and interaction flow. Conversion improved from 17% to 24% and purchase friction dropped.

Protected by NDA

Get in touch to learn more about it.

Product designB2B SaaSAI productsEnterprise UXDesign systemsUX strategyUser researchInteraction designInformation architecturePrototyping
Experience

Where I have done the work.

Product Designer, Paperpal

Cactus Communications
Apr 2024 to Aug 2026
Mumbai / Remote

B2B, Publisher Solutions

  • Manuscript Integrity Screening Platform. Led end to end UX design and scaling for Springer Nature and Taylor & Francis. Designed workflows supporting 2M+ manuscripts and contributed to a 3 year Springer Nature contract (~$600K ARR).
  • AI Manuscript Integrity and Peer Review. Designed AI solutions for the American Medical Association supporting 100K manuscripts a year. Contributed to contracts totaling $150K a year (2 years) and $100K (1 year).
  • Voucher Consumption Dashboard. Designed analytics workflows tracking vouchers issued, consumed and remaining. Surfaced usage trends and upsell and renewal opportunities that informed commercial strategy.
  • Cross functional collaboration. Partnered with Product Managers, Engineering, Data Science, Business and Client stakeholders to define requirements, resolve constraints and drive design through implementation and launch.

B2C, Author Products

  • SmartScreen. Led end to end product design and launch, from discovery and UX strategy through post launch optimization. Scaled to $300K+ ARR and recognized as Best Author Tool at SSP 2026.
  • Human Services Monetization. Designed premium add on experiences for Expert Submission Support, Standard Proofreading and Artwork Formatting. Average order value rose from $28 to $43 (+54%) while premium flow conversion improved.
  • Design and engineering delivery. Created user flows, wireframes, high fidelity UI, interactive prototypes and design specifications. Partnered with Engineering on handoff, implementation reviews, QA and design validation.

Junior UX Designer, Researcher.life

Cactus Communications
Apr 2022 to Mar 2024
Mumbai / Remote

  • Pricing and subscriptions. Designed pricing and subscription experiences for individual and bulk purchases across monthly, quarterly and annual plans, contributing to a 30% increase in conversion.
  • Checkout redesign. Redesigned the multi product checkout journey by simplifying information hierarchy and interaction flow. Conversion improved from 17% to 24%.
  • Cross functional execution. Worked with Product and Engineering to turn requirements into user flows, prototypes and production ready UI.

Education and learning

Bachelor of Commerce with Computer Application, The American College, Madurai, 2017 to 2020

Full Stack Product Management, Airtribe, Jul 2024 to Oct 2024

Full Stack UI/UX Design, Masai, Aug 2021 to Mar 2022

Recognition

3× Design Champion of the Quarter at Paperpal

2× Rising Star of the Quarter at Researcher.life

Best Author Tool, SSP 2026 (SmartScreen)

Writing

Thinking out loud.

See all on Medium
Recommendations

What teammates say.

I had the opportunity to work closely with Mani as part of the same team at Cactus, where I got to see his approach to product design and problem solving firsthand.

Mani has a strong ability to take complex product requirements and translate them into thoughtful, intuitive, and scalable user experiences. His work consistently went beyond creating visually polished interfaces, he paid close attention to user journeys, information hierarchy, interaction patterns, usability, and how each design decision contributed to the larger product objective.

He was particularly effective while working on complex publisher and author facing products, where the challenge was often to simplify large amounts of information and complicated workflows without losing important context. His ability to structure such experiences, think through edge cases, and iterate based on feedback made a noticeable difference to the quality of the product.

From an engineering perspective, I also really valued working with him. Mani collaborated closely with developers throughout the design and implementation process, was receptive to technical constraints, and was always open to discussing alternative approaches when needed. This made the handoff from design to development much smoother and helped ensure that the final product stayed aligned with the original design intent.

What I appreciate most about Mani is that he combines strong design thinking with product awareness. He understands that good design is not only about aesthetics, but about solving the right problem, creating clarity for users, and contributing to measurable product and business outcomes.

He is a thoughtful, collaborative, and dependable Product Designer, and I would confidently recommend him to any team looking for someone who can take ownership of complex product experiences and work effectively across design, product, and engineering.

Manish ChouhanSDE at Cactus Communications, Paperpal PreflightWorked with Mani on the same team

I had the opportunity to manage Mani as a Product Designer and saw him consistently demonstrate strong user understanding and product thinking. He has a natural ability to quickly grasp new domains, understand user needs, and translate that understanding into thoughtful design solutions. He is not only focused on the design outcome, but also on the problem and product context behind it.

Mani is dependable, receptive to feedback, and takes genuine ownership of his work. He collaborates well across teams and brings clarity and thoughtfulness to ambiguous problems. I would confidently recommend him to any product team looking for a strong, growth minded designer.

Sagar KolwankarLead Product Designer, Growth and Activation, PaperpalManaged Mani directly

Let's build something worth shipping.

Open to remote roles and open to relocation. Best reached by email or LinkedIn.

LinkedIn Medium
Mani RajaProduct Designer
← All work
Case study

Paperpal Preflight: catching submission killing issues before an author hits submit

Role

Product Designer, end to end

Company

Cactus Communications

Platform

Embedded web workflow in journal submission systems, plus a Publisher Admin dashboard

Summary

Paperpal Preflight is a free check embedded directly into the submission workflows of major journals, including those of Wolters Kluwer and IOP. Before an author submits, they run the manuscript through Language Quality and Technical Compliance checks, the latter powered by SmartScreen. They get an editorial verdict (Major, Moderate or Minor Revisions Needed) instead of a wall of raw errors.

Publishers configure their own rules per journal through an Admin dashboard, so one product enforces different standards without hardcoded logic. The free report converts into paid AI edited fixes.

The product's success in production later became the internal proof point that gave Springer Nature and the American Medical Association confidence to sign on to a larger enterprise manuscript screening platform.

Context

Editage, Cactus Communications' flagship product, has spent two decades helping researchers get published through manual English editing, scientific review and submission support from real editors. Paperpal Preflight took a slice of that expertise, checking a manuscript before submission, and turned it into a self serve AI product placed at the moment an author is about to submit.

We partnered with journals and publishers, including Wolters Kluwer and IOP, to embed Preflight in their submission flow. SmartScreen, the module that screens for integrity and compliance issues, sits alongside the Language Quality and Technical Compliance checks inside one Preflight report.

IOP Science submission options page for 2D Materials offering a free pre submission check with Paperpal Preflight next to a continue without check option
The free check is offered at the start of a journal's submission flow, next to the option to continue without it.

The problem

Business problem

Journals were losing time and reviewer capacity to a predictable pattern. Manuscripts arrived with fixable issues that surfaced only after an editor or reviewer had already spent time on them. There was no way to catch this earlier without adding friction to submission.

User problem

Authors were not submitting bad research. They were submitting manuscripts that failed rules they did not know existed. Editage's editors, who do this work professionally, told us the real gap was not language quality. It was Technical Compliance: missing conflict of interest statements, missing author contribution statements, and citation placement rules that vary by journal. Most authors did not know Technical Compliance was a category they had to satisfy.

Why now

Initial rejections cost authors months of time and cost journals real review pipeline capacity.

My role

  • Owned end to end design of the author facing Submission Readiness Report and the flag and status system underneath it
  • Owned design of the Publisher Admin dashboard for configuring journal specific checks
  • Owned the monetization flow, from free diagnostic to paid AI edited fix
  • Worked closely with Editage's professional editors and with the ML team training the models on 20 years of published paper data
  • Did not own the NLP and ML model architecture, or publisher contract negotiation

Discovery

We had an advantage most AI products do not get: direct access to the people who did this work by hand before any of it was automated.

  • Editor interviews. We sat with Editage's editors, who are trained across manuscript types and journal specific nuances, to learn what each journal actually required. This is where we found the core insight: manuscripts rarely failed on language. They failed on Technical Compliance requirements authors had never heard of.
  • 20 years of published paper data. We trained the system on what correct looks like across article types and journals, instead of applying one generic standard.
  • Per journal templates. We built configurable checklists per journal and article type, because a rule that is critical for one journal, such as a strict abstract word count, can be irrelevant for another.

Designing for AI trust and uncertainty

This product had a specific trust problem. It told authors who were not experts that they had failed rules they did not know existed, using an AI system, with no human editor to appeal to. Three decisions shaped how we handled it.

1. Give journals control instead of one hardcoded standard

We built a Publisher Admin dashboard where journal staff configure their own checklist, including Abstract Length, List of Contributing Authors, Author Contributions Statement, citation rules, and figure and table requirements, and mark any of them as critical. Pass and fail results roll up into one of three plain language statuses: Major Revisions Needed, Moderate Revisions Needed or Minor Revisions Needed. Without this, a check that is critical for one journal would wrongly block or wrongly pass a manuscript for a journal with different standards.

Paperpal Preflight Admin screen for article type configuration showing checks such as Abstract Length and List of Contributing Authors, each with a Mark as CRITICAL option
Publisher Admin: configuring the checks for an article type, with each check markable as CRITICAL.

2. Structure the report so it teaches, not just accuses

The report leads with the verdict, for example "Major Revision Required", then splits into Language Quality and Technical Compliance. An author can tell at once whether the issue is something they can simply fix (writing) or a rule they have never met (compliance). Within Technical Compliance, critical failures look different from minor ones, and passed checks are shown too. Since most authors did not know these rules existed, showing what they got right next to what they got wrong was deliberate. It frames the report as guidance, not a rejection letter.

Submission Readiness Report with a Major Revision Required summary, Language Quality and Technical Compliance counts, and an unlock panel for the AI edit and review
The report at Major Revision Required. The summary leads, then Language Quality and Technical Compliance.
Submission Readiness Report with a Minor Revisions Needed summary and a panel saying the AI edited manuscript is ready
The same report at Minor Revisions Needed.
Technical Compliance section grouped into Counts, References and Disclosures with failed checks expanded and passed checks collapsed
Technical Compliance grouped by Counts, References and Disclosures, with passed checks collapsed.
Detailed report listing failed checks with red markers, warning triangles on critical ones, and green passed groups
Failures and passes side by side, with critical failures marked.

3. No dispute flow, unlimited free rechecks instead

Authors cannot appeal or ask why on a specific flag. They can fix the issue and rerun the check as many times as they want, for free, before submitting. This was a deliberate tradeoff. Because it happens before submission with no real time stakes, the system does not need to be right on appeal. It needs to be cheap and fast to verify again. That let us skip a moderation or escalation queue entirely, at the cost of authors having no way to contest a flag they believe is wrong.

Key design decisions

Give the diagnosis away free, charge for the fix

The Submission Readiness Report, including status, issue counts and critical flags, is free to every author who arrives through a partner journal. The paid tier ($29 to $49) unlocks an AI edited manuscript with fixes applied and annotated, plus a deeper compliance and methodology report.

Why. The free report proves value at the moment of highest intent, when an author is about to submit and worried about rejection. Charging only for the fix turns trust already earned into revenue, instead of asking an author to pay before seeing what is wrong.

Tradeoff. Revenue depends on the free tier being good enough to build trust, but not so complete that authors have no reason to buy the edited file.

AI edit and review purchase panel showing the price, an output preview of tracked fixes and what the purchase includes
The premium panel previews the AI edited output before the author pays.
Word document with Paperpal tracked changes and margin comments on a sample manuscript
The paid output: an AI edited manuscript in Word with tracked changes and comments.

Roll granular checks up into three plain language statuses

Instead of a flat pass or fail list, we translated many individual checks into one editorial verdict (Major, Moderate or Minor Revisions Needed), the same language editors use.

Why. Authors do not think in terms of 15 checks failed and 11 passed. They think: am I ready to submit or not?

Tradeoff. It required getting the weighting of critical and non critical checks right per journal. Get it wrong and the top line verdict misleads even when the detail underneath is accurate.

Outcomes

MetricResult
Product recognitionBest Author Tool, SSP 2026
RevenueScaled to $300K+ ARR

Preflight's success in production proved that the underlying screening AI worked reliably at scale across real journals. It became the internal case for pitching enterprise publisher deals, and is part of why Springer Nature and the American Medical Association had the confidence to sign on to the larger enterprise manuscript screening platform that followed.

Reflection

What I would do differently

I would test whether a lightweight "flag this as wrong" signal, short of a full dispute or appeal flow, could catch systematic errors in specific journal configurations faster, without building a full moderation queue.

What surprised me

How much of the value was in surfacing a hidden rulebook instead of fixing writing quality. The product's real job was not to make a manuscript well written. It was to tell authors the rules they did not know they had to follow.

What I carried forward

Publisher side configurability, the Admin dashboard, was the unglamorous but essential piece that made one product credible across dozens of journals with different standards. The same instinct, make the standard configurable and not hardcoded, shaped how I approached the enterprise manuscript screening platform that followed.