What is Navin Sharma's engineering background?

Founder of GetWidget (legal entity: Paiteq Private Limited) and operator since 2017. Daily Claude Code and OpenAI Codex operator. Eight years on the open-source GetWidget Flutter UI Kit (4,811 GitHub stars, 23,000 monthly pub.dev downloads) before pivoting the studio to production AI delivery in 2024.

What kinds of projects does Navin lead?

Production AI workflows in 4 to 9 weeks: RAG pipelines, agent orchestration with LangGraph, voice agents on the OpenAI Realtime API, and intelligent document processing. Regulated-buyer deployments under HIPAA, FFIEC, SR 11-7, and ISO 42001. Flutter plus AI in the same team.

What does Navin not take on?

50+ headcount ML team work (Big-4 territory), single-vendor SaaS resale, pure research without a workflow target, demos without measurable outcomes, and single-model vendor lock-in. Model-agnostic by default across Claude, OpenAI, and open-source.

How do I book an audit with Navin?

Start the audit conversation on the contact page. The 30-minute intro is free; the full discovery audit is a fixed-scope, 1-2 week engagement that ends with a written prioritisation, a draft eval rubric, and a go/no-go recommendation.

Founder · GetWidget

Navin Sharma

Founder of GetWidget (legal entity: Paiteq Private Limited). Operator since 2017. Ships production AI for regulated healthcare + fintech + legal buyers.

What I do

I run GetWidget, a Bengaluru- and Dallas-based AI development studio. Daily Claude Code and OpenAI Codex operator. We ship production AI agents, RAG systems, voice agents, and intelligent document processing for clients across 10 industries. Eval-first delivery, model-agnostic across Claude + OpenAI + open-source. Audit-logged, regulator-defensible. Engagements run as a discovery audit, then a 4-6 week pilot with weekly eval gates, then continuous delivery in production. Our pilot RAG run logged 88% faithfulness on a 1,840-document corpus (2026-Q1); the agent harness logged 71% pass@1 across 100 tool-using tasks (2026-Q1).

How GetWidget started

In 2017 we began building a Flutter UI kit because Flutter components were inconsistent across our client projects. Eight years later the kit sits at 1,000+ widgets, 4,811 GitHub stars, and 23,000 monthly downloads on pub.dev, powering production mobile apps for teams across the US, India, and Europe.

In 2024 we pivoted the studio to AI. Same engineering team, same discipline, different problem. Today GetWidget ships production AI for clients across 10 industries. The OSS kit is the reputation backstop; the AI work is where the team's day-to-day energy goes.

What I write about

Long-form essays on AI delivery, LLM evaluation methodology, RAG architecture, voice-agent latency budgets, agent orchestration with LangGraph and CrewAI, and regulated-buyer deployment under HIPAA, SR 11-7, FFIEC, and ISO 42001.

Recent posts live at /blog/author/navin/. The flagship operator-grade reference is AI Governance Readiness 2026.

What I will and won't take on

Yes: production AI workflows in 4 to 9 weeks with named eval criteria, regulated-buyer deployments (HIPAA, FFIEC, SR 11-7), model-agnostic engineering across Claude, OpenAI, and open-source, Flutter plus AI in the same team.

No: 50+ headcount ML team work (Big-4 territory), single-vendor SaaS resale, pure research, demos without measurable workflow value, single-model vendor lock-in.

Where to find me

LinkedIn: linkedin.com/in/navinsharma002
X / Twitter: x.com/navinsharmacse
GetWidget OSS repo: github.com/ionicfirebaseapp/getwidget
Audit booking: contact page (30-minute free intro or fixed-fee full discovery audit)