Rutvik Makwana
--:--

Human Data Team Lead · xAI

I build teams and systems that keep LLM training sharp and boringly reliable.

2+ years working on Grok models. Led a 24‑person annotation team in IST, owned hiring, training, and QA. I care about clean guidelines, honest metrics, and fixing problems at the process level.

Contact LinkedIn
24 people led
Grok 1.5 · 2 · 3
QA hundreds of tasks / wk

What I actually did

Short version of the xAI story.

At xAI I went from AI Tutor to Human Data Team Lead. I ran an IST‑based team of 24 people working on Grok training: reviewing multi‑modal tasks, catching failure patterns early, and keeping standards consistent across projects.

I designed onboarding flows, wrote training material, interviewed candidates, and kept performance visible using Axiom and internal tools. Most of my time went into making sure “good work” meant the same thing for everyone on the team.

How I think about the work

Not buzzwords. Just how I run a team.

Quality

Small, frequent reviews. Clear examples. Shared edge‑case playbook.

People

Hiring screens, honest feedback, and 1:1s that actually talk about work.

Process

If the same mistake shows up twice, the guideline or workflow changes.

Why the medical background matters

You can trust me with high‑stakes domains.

Pharmacy & trials

B.Pharm with pharmacology and pharmaceutics. M.Pharm in Regulatory Affairs with coursework in clinical trials and regulatory writing.

Regulatory work

Thesis drafting an ANDA for USFDA approval and writing about USFDA / EMA processes for Regulatoryaffairs.education.

Contact

If you have a serious AI training problem, send context.