Prakash Adhikari
Contact

About

A builder journey, not a résumé.

From systems work in 2018 to intelligent software used by organizations — the path that shaped how I design products.

Portrait of Prakash Adhikari

I build software that desks depend on — SaaS platforms, automation, and domain tools where failure shows up as a missed renewal, a lost inquiry, or a client call that should never have been needed.

Today that practice is AI product building: document intelligence, retrieval, and systems that amplify expert judgment. Based in Kathmandu. Shipping for real organizations.

AI Product Builder Software Architect Systems Engineer

Career narrative

How the work got here.

2018Started building systems — infrastructure, support, and the operational backbone organizations already depended on.
Internal automationScripts and tools that removed repeated pain. First lesson: useful software disappears into the workday.
Enterprise workflow softwareClient platforms, education products, then LegalTech — software shaped by how practitioners actually file, search, and renew.
AI-powered productsOCR pipelines, register-scale search, retrieval-first features with humans in the loop — inside production, not pitch decks.
TodayBuilding intelligent software systems — AI product development, architecture, and products organizations can trust.

What this path taught

Reliability is a product feature.

Teaching computer science sharpened how I explain systems. Shipping NepalIPMS sharpened how I choose stacks. Sitting with IP desks sharpened what “done” means.

  • Systems firstUnderstand the desk before choosing the framework.
  • Production over demosIf it does not survive Monday morning, it is not shipped.
  • AI with guardrailsExtraction and drafts accelerate experts — they do not replace judgment.
  • Small-team edgeCloudflare-native stacks when ops overhead would kill the product.

The person

Enough personality to remember.

  • Principle I changedFuzzy-first search → exact-first trust after production ambiguity on the trademark register. Story →
  • Tools I reach forNotebook before Figma · TypeScript · Workers / D1 / R2 · a quiet terminal
  • Failed project lessonBroad similarity search looked smart in demos and failed during live client calls — trust is the feature.
  • TeachingExplaining systems to students forces the same clarity products need
  • PortraitEditorial photos coming — photographer brief · framed slot ready on this page

How I build → · Artifacts →

Organizations

Work that runs in production.

Verified products and known operating contexts — no fabricated testimonials.

  • Global Law AssociatesWebsite/CMS; IP desk on NepalIPMS.
  • Janak Bhandari & AssociatesNepalIPMS for trademark and IP matter management.
  • ScholarQuest · Sri · TopRank · YogshalaLive education and wellness platforms in Nepal.
  • TeachingComputer science, web technologies, applied product thinking.

Available for

Let’s talk about your system.

  • AI Product Development
  • Technical Architecture
  • SaaS Consulting
  • Automation Systems
  • Product Strategy
hello@prakashadhikari.dev