AI Builder

Khush

Agarwala.

~/portfolio $ whoami

khush agarwala, AI builder

~/portfolio $ ls apps/

6 live apps, all running ✓

~/portfolio $ cat status.txt

building with AI ● shipping products

~/portfolio $

View my work ↓ GitHub

scroll

LLM APIs· FastAPI· SSE Streaming· LLM-as-Judge· Parallel Evals· Web Speech API· GitHub API· SQLite· Pydantic v2· Tool Use· Structured Extraction· Python· Vanilla JS· CSS Animations· Async/Await· LLM APIs· FastAPI· SSE Streaming· LLM-as-Judge· Parallel Evals· Web Speech API· GitHub API· SQLite· Pydantic v2· Tool Use· Structured Extraction· Python· Vanilla JS· CSS Animations· Async/Await·

Projects

Things I've
been building.

A handful of AI tools I put together. Click around, they actually work.

gtm-intelligence — live

GTM Intelligence Suite

Full pre-call brief in one shot. Paste a company + contact and get a company snapshot, personalized outreach email, call script, objection handling, and recommended next step. The entire GTM research workflow, collapsed.

WORKFLOW AUTOMATIONSTREAMINGMULTI-SECTION OUTPUTGTM

live ↗ Open app →

pr-review — live

PR Diff Review

Paste a GitHub PR diff and get an instant AI code review with inline comments, risk flags, and suggested changes.

CODE REVIEWCLAUDE API

live ↗ Open app →

eval-runner — live

Eval Runner

Run structured LLM evals against test cases, score outputs, and surface regressions. Built for AI product teams.

EVALSLLM TESTING

live ↗ Open app →

doc-extractor — live

Doc Extractor

Upload any document and extract structured data, tables, key fields, and summaries. Handles PDFs, images, and messy formats.

DOCUMENT AIEXTRACTION

live ↗ Open app →

Khush Agarwala.

Things I'vebeen building.

Things I'm thinking about.

Khush

Agarwala.

Things I've
been building.