Pivony — AI Engineer (Contractor, LLM & Applied ML) at Pivony
What is Pivony?
Pivony is a customer intelligence platform that helps teams turn large-scale customer feedback (reviews, surveys, social, forums) into clear insights and actions. We combine analytics dashboards with AI to help organizations understand what customers love, what’s broken, and what to do next.
Why join us?
We’re a small, compact team that ships fast and values ownership. Contractors work close to the product and decision-makers—less bureaucracy, more impact. If the collaboration goes well, there’s a path to long-term work and internal progression.
What you’ll do
Build and ship AI/LLM capabilities used in production workflows (evaluation-driven).
Deploy and operate models responsibly (latency, cost, reliability, monitoring).
Work with backend/frontend to integrate AI outputs into real product features.
How we work
We pay based on accepted deliverables/outcomes, not hours.
We value measurable results: evaluation, regression control, and reproducibility.
Requirements
1+ year of professional experience in applied ML/LLM engineering (or equivalent proven track record).
Must have deployed LLMs on-prem (self-hosted / private infrastructure): setup, serving, and operations.
Strong understanding of evaluation (offline metrics, acceptance criteria, regression testing).
Able to join video calls reliably (good bandwidth, clear audio/video).
Nice to have (non-exclusionary)
Strong academic record and rigor (e.g., high GPA, awards, publications) is a plus, but equivalent industry impact is equally valued.
Experience with RAG, embeddings, vector databases, and prompt/model iteration loops.
Experience optimizing inference (quantization, batching, caching, GPU utilization).