Senior LLM Evaluation & Fine-Tuning Architect

Innodata Inc.; United States; On-site; USD 140,000 - 160,000; Full time

Job summary

Innodata in the United States is seeking a Senior Technical Solutions Architect for Evals & Fine-Tuning to be the technical face to demanding customers. You will translate objectives into scoped engagements and lead technical discovery with foundation model teams and enterprise AI groups.

The role requires hands-on experience in fine-tuning LLMs, designing eval harnesses, and credible communication with researchers and executives.

Qualifications

• 7+ years in applied ML, ML engineering, ML research, or technical solutions.
• 2+ years focused on LLM evaluation and/or post-training.
• Hands-on fine-tuning of LLMs (SFT; RLHF/DPO preferred).
• Deep familiarity with LLM evaluation, benchmarks, and human eval workflows.
• Strong Python and modern LLM toolchain knowledge.
• Excellent technical communication and stakeholder management.
• Bachelor’s or advanced degree in CS/ML or equivalent experience.

Responsibilities

• Lead technical discovery with foundation model labs and large enterprises to understand model objectives and constraints.
• Design end-to-end post-training solutions: SFT data curation, RLHF/DPO data, golden datasets, benchmarks, LLM‑as‑judge pipelines, and evaluation harnesses.
• Architect engagements combining GenAI platforms with a global SME workforce across 85+ languages and domains.
• Author proposals, SOWs, diagrams, and pricing in collaboration with sales, delivery, and finance.
• Run workshops, POCs, and pilot designs to de-risk larger programs and prove value.
• Act as ongoing technical advisor during delivery with researchers, ML engineers, and program managers.
• Feed customer signal back into R&D and product roadmap; monitor eval methodologies.
• Stay current on state‑of‑the‑art evals and post-training approaches.

Skills

LLM evaluation

Python

ML engineering

Technical communication

Consultative mindset

SFT data curation

RLHF/DPO

Education

Bachelor’s or advanced degree in computer science / ML / computational linguistics

Tools

Hugging Face

PyTorch

vLLM

lm-evaluation-harness

Job description

Innodata in the United States is seeking a Senior Technical Solutions Architect for Evals & Fine-Tuning to be the technical face to demanding customers. You will translate objectives into scoped engagements and lead technical discovery with foundation model teams and enterprise AI groups.

The role requires hands-on experience in fine-tuning LLMs, designing eval harnesses, and credible communication with researchers and executives.