Nelson Sarma

Hello.

I am Nelson Sarma. I look after twelve designers across five crafts that agree on almost nothing, which is most of the fun. My job is no longer to be the person with all the answers. It is to make sure the right person can make the call, has enough context to make it well, and has someone behind them when it gets messy. I am still in the files most weeks.

Nelson Sarma, arms folded.
  • 7yrs In
    Design
  • 2.5yrs In
    Leadership
  • 12 People on
    the team
  • 5 Design
    Domains

From designer to managing designers.

I started by learning how to make good work. Then I learned how to create the conditions for other people to make it.

  1. 01

    Graphic designer

    I learned to make.

    • Brand
    • Campaigns
    • Visual communication
    • Visual craft
    • Composition
    • Typography
    • Image

  2. 02

    UI designer

    I learned to design for screens.

    • Websites
    • Interfaces
    • Interaction
    • Design systems
    • Interface
    • System
    • Interaction

  3. 03

    Design manager

    I learned to make space.

    • People
    • Ownership
    • Systems
    • Operations
    • Team
    • Environment
    • Operating model

  4. Today

    • 12 designers
    • 5 crafts
    • 2.5+ years leading

I didn’t leave the craft when I took the title.

From Guwahati, Assam. Design was the passion before it was the job: seven years in, the last two and a half leading twelve designers across five crafts, from a graphic designer to a web designer and lead. Management didn’t make design less important to me; it changed where I could have the most impact. I stopped measuring my contribution by how many screens I personally ship, and started measuring it by what the team can ship without me becoming the bottleneck. I still open the file, still write the bar down, and still take the work into the codebase behind engineering’s guardrails, because that is usually the fastest way to understand what I am asking people to build.

  • Five crafts, one bar

    Product, social, web, presentation, illustration. All under one umbrella, held to one written standard, so “good” doesn’t depend on who happened to review it.

  • Into the frontend

    Design systems, HTML and SCSS, shipped into the product. It keeps me honest about what I design.

  • Agents take the tedious half

    First drafts and scaffolding go to the machines. Judgement stays with people.

Most design problems aren’t design problems.

A designer can be very good at their craft and still struggle when ownership is unclear, priorities keep moving, feedback arrives from five directions, or every decision has to route through a manager. So I spend a lot of my time fixing the environment around the work. Six frameworks, each built the week something stopped working. None is revolutionary on its own; together they let designers work with real autonomy without being left alone when it gets difficult:

  1. I turned a Git repo into a team OS.

    Work, OKRs, availability, AI skills, assets and team history live in one shared workspace, which makes knowledge and work transferable across the team. The team stopped waiting for people.

  2. Everyone has a backup. Nobody staffs the third layer.

    Assigning designers to domains is what every team does. Staffing the layer under the backup is not. The goal isn’t perfect coverage; it is making sure one person’s absence never becomes everyone else’s emergency.

  3. Intake, and the polite art of moving a date.

    If everything is urgent, nothing is planned. Six fields and published turnaround times, so an estimate is agreed once rather than renegotiated every time, and three routes for when a date genuinely has to move. Built to protect designers’ time, not to control it.

  4. Velocity and peer review, running continuously.

    Two angles: what shipped, and how it was to work with. Taste stays out of it, the team sets the weights rather than me, and judgement stops being something only the manager is qualified to sign off.

  5. The annual review as reconciliation.

    A year of data we both already have. No need to perform in the fourth quarter, since it counts for exactly as much as the first.

  6. I went first, and got audited hardest.

    I negotiated the guardrails with engineering before anyone had access, then ran the deepest adoption in the company myself. The goal was never to make twelve designers use AI. It was to make sure they knew when not to.

Hands-on

Still building.

Recent work across systems, accessibility and product experience, shaped with the same hands-on curiosity that got me into design.