Senior Artificial Intelligence/Machine Learning Engineer
New
J
JobgetherDocument AI
Remote position in India.Full-TimeSenior
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Required Skills
- AWSPrompt Engineering
Requirements
- Deep experience in AWS architecture, including hands-on production work with Amazon Bedrock.
- Strong expertise in model selection, prompt engineering, context engineering, and model evaluation.
- Experience optimizing inference costs and latency.
- Substantial document AI and OCR experience, including degraded-input processing with Amazon Textract or equivalent technologies.
- Understanding of layout-aware extraction and converting OCR output into reliable, schema-compliant structured data.
- Experience extracting information from long, unstructured documents into fixed schemas, including field-level confidence estimation and calibration.
- Experience building agentic AI systems with multi-step reasoning, tool-use protocols, and agent state management.
- Production experience with RAG or LLM pipelines, including document chunking, retrieval strategies, contextual grounding, and knowledge integration.
- Experience designing human-in-the-loop workflows that route uncertain outputs for review.
- Ability to incorporate reviewer corrections into continuous-improvement feedback loops.
- End-to-end ownership of ingestion, inference, evaluation, monitoring, and production feedback mechanisms.
- Strong problem-solving, communication, and collaboration skills, including the ability to guide engineers and establish engineering practices.
Responsibilities
- Design, build, test, and maintain production extraction agents that convert unstructured documents into structured schemas.
- Develop ingestion pipelines for degraded scans, photocopies, mixed image-and-text PDFs, and inconsistent layouts.
- Build field-level confidence models and route uncertain outputs to human review.
- Configure RAG knowledge bases that provide secure, domain-specific context and improve grounding.
- Implement review and escalation workflows, confidence thresholds, and output controls.
- Capture reviewer corrections as feedback to improve future system performance.
- Develop workflows that compare new extraction results with existing records and highlight meaningful changes.
- Optimize model selection, prompts, context, token use, inference cost, and latency.
- Establish evaluation and monitoring mechanisms to assess accuracy, production performance, and reliability.
- Own the AI system lifecycle and mentor engineers on reusable extraction, confidence, orchestration, and system-design patterns.
View Full Description & ApplyYou'll be redirected to the employer's site
