AI Engineer  

Job Desk

Design and develop AI microservices architecture for OCR, LLM services, embedding services, and Retrieval-Augmented Generation (RAG) pipelines.

Manage AI system deployments on Kubernetes (AWS EKS) using GPU-enabled nodes (A10G, T4, G5), Helm, and Terraform.

Configure and maintain MLOps infrastructure, including IAM/IRSA, EBS/EFS CSI, and monitoring solutions such as Grafana, Loki, and DCGM Exporter.

Optimize AI inference performance using ONNX, TensorRT, OpenVINO, and vLLM through batching, concurrency tuning, and prefix caching.

Analyze and improve throughput, latency, GPU memory utilization, and KV cache efficiency for LLM and OCR services.

Design and execute load testing and benchmarking for AI services using Locust, k6, and custom testing tools.

Prepare technical performance testing reports for technical stakeholders and management.

Develop and maintain a microservices-based OCR system for large-scale document processing.

Build modular OCR pipelines, including document detection, orientation correction, text detection, and text recognition.

Manage the conversion and optimization of OCR models into production-ready inference formats such as ONNX and TensorRT.

Design and implement OCR APIs with support for batching, concurrency, and dynamic parameter configuration.

Optimize image preprocessing and post processing pipelines to improve OCR accuracy and system stability.

Develop Multimodal RAG and Visual Question Answering (VQA) systems using ColBERT/ColPali and Vision-Language Models (VLMs).

Build embedding and retrieval systems for visual documents based on text, layout, and image features.

Develop and maintain the LLM service stack, including FastAPI, Redis, Kafka/SQS, and object storage.

Design dedicated logging and analytics pipelines to support auditing and performance analysis.

Conduct experiments involving LoRA fine-tuning, quantization, and model distillation to improve inference efficiency.

Investigate CUDA determinism issues, floating-point drift, and numerical stability in production AI systems.

Develop and maintain technical documentation, deployment standard operating procedures (SOPs), operational runbooks, and load testing reports.

Have exp 3 years as AI Engineer

Prepare the AI model integration plan.

Integrate the AI model into BRI's systems.

Validate the AI model integration results.

Prefer has exp in Banking industry