Senior AI Engineer: Remote LLM Infra for Docs

Aadvik Tech Services Pvt Ltd; Northern (KY); Hybrid; USD 140,000 - 200,000; Full time

Benefits offered by this job

Remote work (US shift)

Job summary

Aadvik Tech Services Pvt Ltd in the United States is seeking a Senior AI Engineer to own and scale OCR-driven document intelligence pipelines and self-hosted LLM infrastructure. You will build production-grade systems that process long-form documents with deterministic extraction.

The role demands 6+ years of Python experience, hands-on work with vLLM/Hugging Face TGI, layout-aware models, GPU-backed serving, and robust validation layers.

Qualifications

• 6+ years of Python engineering experience.
• Production experience building OCR-driven pipelines.
• Experience with long-form PDFs (100+ pages).
• Experience with vLLM or Hugging Face TGI and GPU-based serving.
• Experience building deterministic validation systems (schema + rule enforcement).
• Strong debugging and systems-level thinking.

Responsibilities

• Design end-to-end pipelines for processing long-form OCR-heavy documents.
• Own PDF ingestion, layout-aware parsing, and multi-page assembly.
• Implement deterministic schema validation and cross-field checks.
• Deploy and operate open-source LLMs with GPU-backed serving stacks.
• Tune performance: KV cache, batching, context window, latency vs throughput.
• Ensure auditability and reproducibility of extraction results.

Skills

Python

OCR pipelines

Long-form PDFs

vLLM / HF TGI

GPU serving

Deterministic validation

System design

Tools

vLLM

Hugging Face TGI

Open-source LLMs

Job description

Aadvik Tech Services Pvt Ltd in the United States is seeking a Senior AI Engineer to own and scale OCR-driven document intelligence pipelines and self-hosted LLM infrastructure. You will build production-grade systems that process long-form documents with deterministic extraction.

The role demands 6+ years of Python experience, hands-on work with vLLM/Hugging Face TGI, layout-aware models, GPU-backed serving, and robust validation layers.