Arta Ghasemi · Engineering

I build practical products and the launch loops around them.

CS @ Penn State. Previously shipped RAG pipelines as a SWE intern. Currently building Test Maxer, a men's health optimization app, and handframe.app, a paywalled file-delivery platform for freelancers.

Founder Signals

CS GPA

3.92

Penn State Computer Science while shipping products and client work.

first Swift app

3-4 mo

Learned, built, and shipped Test Maxer end-to-end.

downloads

200+

Test Maxer traction while testing the next product decisions.

launch posts

60

Product marketing shipped alongside the app, not after.

Selected Work

• Test Maxer

  Personal iOS product · validation

  Men's health optimization iOS app

  Built and shipped a consumer app for men interested in natural testosterone optimization. Includes AI-powered face scanning, habit tracking, personalized action plans, and a scoring loop.

  Outcome

  200+ downloads, 60 short-form launch posts, and an active user-feedback sprint before the next product decision.

  Stack

  Swift · iOS · AI vision API · App Store

  Read case study →

• handframe.app

  Personal product · shipping

  Paywalled file delivery for freelancers

  Platform for freelancers to create a paywalled download link for final files. Clients pay before accessing the video or asset, reducing late payment and post-download ghosting.

  Outcome

  Built around the handoff moment where freelance work gets approved, paid for, and downloaded.

  Stack

  Web app · Payment links · File delivery

  Read case study →

• RAG pipeline

  SWE internship · 2025

  Retrieval infra for a consumer app

  Shipped a retrieval pipeline with 4 strategies behind a shared evaluator. Built the eval harness, A/B rollout, and telemetry dashboards.

  Outcome

  35% lower eval latency, rolled out to weekly active users.

  Stack

  Python · pgvector · OpenAI · Weights & Biases

  Read case study →

• Validation sprintDirect outreach to Test Maxer users before the next build decision.
• Penn State courseworkProjects from CMPSC 360, 465, 442.
• Build logShort dev notes on AI tooling and evals.

Experience

1. SWE Intern

   Stealth consumer AI startup

   Summer 2025
   • Built RAG pipeline reducing eval latency 35% across 4 retrieval strategies.
   • Shipped telemetry dashboards surfacing per-query retrieval hit rate.
   • Owned the eval harness; wired it into CI for regression gating.

   Python · pgvector · OpenAI · Vercel

2. CS Student

   Penn State — College of Engineering

   2024 — Present
   • Maintaining a 3.92 GPA while building products and running freelance video work.
   • Currently focused on applied ML, systems, data structures, and consumer app development.

   Python · TypeScript · C · Go

Writing

• Notes on shipping Test MaxerDraft
• What a real RAG eval loop looks likeDraft
• Small models, real productsDraft