HiQo Solutions, Inc.
Fullstack AI/ML Developer (office/remote)
HiQo Solutions, Inc., Poland, New York, United States
Overview
We're looking for a versatile AI/ML Specialist (Generative & Multimodal) who can translate business problems into data-driven solutions and deliver models to production in close collaboration with our team. You'll work across data exploration, modeling, and integration, with a strong emphasis on modern AI from classical machine learning to generative models and LLMs.
In this role, you'll build and operate end-to-end ML systems, owning data pipelines, modeling, deployment, monitoring, and iteration. The ideal candidate combines solid software engineering fundamentals with practical machine learning expertise and hands-on experience in MLOps and cloud environments.
Location:
Poland (office in Krakow/Wroclaw or 100% remote)
RESPONSIBILITIES
Translate business problems into ML/AI solutions and measurable success criteria.
Build reliable data pipelines (batch/stream) for training and inference; implement data validation and quality checks.
Develop, train, and evaluate models (classical ML and deep learning) with reproducible experiments.
Design and ship production services (APIs, batch jobs, streaming consumers) with automated tests and observability.
Establish and maintain MLOps foundations: versioning (code/data/models), experiment tracking, model registry, CI/CD, and automated deployments.
Monitor production systems (latency, throughput, cost, model performance, drift) and implement retraining/rollbacks.
Apply modern AI techniques: LLM integrations including multimodal, retrieval-augmented generation, visual embeddings for RAG, prompt design and evaluation, guardrails.
Optimize cost and performance (profiling, batching, caching, quantization, GPU utilization) and ensure reliability.
Collaborate with product, data, and engineering stakeholders; document designs and decisions.
REQUIREMENTS
3-5+ years building ML-powered products with production ownership (data - model - deployment - monitoring).
Strong Python and software engineering fundamentals: clean code, testing, logging, type hints, code reviews, modular design.
Proficiency with ML/DL stack: scikit-learn; PyTorch or TensorFlow; pandas/NumPy; solid grasp of evaluation metrics and experiment design.
SQL and data modeling; experience with warehouses/lakehouses (e.g., BigQuery/Snowflake/Redshift) and ETL/ELT tools.
Production experience with image/video ML: classification, object detection, segmentation, tracking, OCR/document AI.
CV frameworks and libraries: OpenCV; PyTorch (torchvision, timm) or TensorFlow/Keras; YOLO/Detectron2/MMDetection; augmentation with Albumentations.
Orchestration and pipelines: Airflow/Prefect/Dagster or similar.
Containers and deployment: Docker; basic Kubernetes or serverless; API frameworks (FastAPI/Flask).
Cloud experience (AWS/GCP/Azure) including storage, compute, networking, and IAM basics.
MLOps tooling: experiment tracking and model management (MLflow, Weights & Biases), model registry, artifact/version control.
Monitoring/observability: metrics, tracing, and alerting (Prometheus/Grafana/CloudWatch/Datadog); model drift monitoring.
Practical AI/LLM experience: using hosted APIs or open-source models, embeddings/vector databases (FAISS/Pinecone/pgvector), RAG patterns, safety/guardrails.
Clear communication and the ability to scope, estimate, and deliver incrementally.
Education: BS/MS in Computer Science, Data Science, Statistics, Engineering or equivalent practical experience.
English Proficiency at least B2 level.
NICE TO HAVE
Infrastructure as Code (Terraform/CloudFormation), Helm, KServe/SageMaker/Vertex AI/Azure ML.
Streaming systems (Kafka/Kinesis/Pub/Sub) and real-time inference.
Feature stores, data contracts, and data governance.
Multimodal and generative vision: CLIP/BLIP/ALBEF, Vision Transformers, diffusion models (inpainting, super-resolution), visual RAG.
Performance tuning: ONNX/TensorRT, quantization, distillation, GPU scheduling; Numba/Cython.
Security and compliance basics for ML systems, PII handling, secrets management.
A/B testing, causal inference, and product analytics.
WE OFFER YOU
Flexible working time - you can agree on it within the team
Necessary tools and equipment
Communication in English - only foreign customers, and international teams
Simple structure and 'open door' way of communication
Full-time English teachers
Medical insurance for employees
HiQo University- internal education and training programs
HIQO COINS - We have a system of rewarding employees for extracurricular activities
#J-18808-Ljbffr
In this role, you'll build and operate end-to-end ML systems, owning data pipelines, modeling, deployment, monitoring, and iteration. The ideal candidate combines solid software engineering fundamentals with practical machine learning expertise and hands-on experience in MLOps and cloud environments.
Location:
Poland (office in Krakow/Wroclaw or 100% remote)
RESPONSIBILITIES
Translate business problems into ML/AI solutions and measurable success criteria.
Build reliable data pipelines (batch/stream) for training and inference; implement data validation and quality checks.
Develop, train, and evaluate models (classical ML and deep learning) with reproducible experiments.
Design and ship production services (APIs, batch jobs, streaming consumers) with automated tests and observability.
Establish and maintain MLOps foundations: versioning (code/data/models), experiment tracking, model registry, CI/CD, and automated deployments.
Monitor production systems (latency, throughput, cost, model performance, drift) and implement retraining/rollbacks.
Apply modern AI techniques: LLM integrations including multimodal, retrieval-augmented generation, visual embeddings for RAG, prompt design and evaluation, guardrails.
Optimize cost and performance (profiling, batching, caching, quantization, GPU utilization) and ensure reliability.
Collaborate with product, data, and engineering stakeholders; document designs and decisions.
REQUIREMENTS
3-5+ years building ML-powered products with production ownership (data - model - deployment - monitoring).
Strong Python and software engineering fundamentals: clean code, testing, logging, type hints, code reviews, modular design.
Proficiency with ML/DL stack: scikit-learn; PyTorch or TensorFlow; pandas/NumPy; solid grasp of evaluation metrics and experiment design.
SQL and data modeling; experience with warehouses/lakehouses (e.g., BigQuery/Snowflake/Redshift) and ETL/ELT tools.
Production experience with image/video ML: classification, object detection, segmentation, tracking, OCR/document AI.
CV frameworks and libraries: OpenCV; PyTorch (torchvision, timm) or TensorFlow/Keras; YOLO/Detectron2/MMDetection; augmentation with Albumentations.
Orchestration and pipelines: Airflow/Prefect/Dagster or similar.
Containers and deployment: Docker; basic Kubernetes or serverless; API frameworks (FastAPI/Flask).
Cloud experience (AWS/GCP/Azure) including storage, compute, networking, and IAM basics.
MLOps tooling: experiment tracking and model management (MLflow, Weights & Biases), model registry, artifact/version control.
Monitoring/observability: metrics, tracing, and alerting (Prometheus/Grafana/CloudWatch/Datadog); model drift monitoring.
Practical AI/LLM experience: using hosted APIs or open-source models, embeddings/vector databases (FAISS/Pinecone/pgvector), RAG patterns, safety/guardrails.
Clear communication and the ability to scope, estimate, and deliver incrementally.
Education: BS/MS in Computer Science, Data Science, Statistics, Engineering or equivalent practical experience.
English Proficiency at least B2 level.
NICE TO HAVE
Infrastructure as Code (Terraform/CloudFormation), Helm, KServe/SageMaker/Vertex AI/Azure ML.
Streaming systems (Kafka/Kinesis/Pub/Sub) and real-time inference.
Feature stores, data contracts, and data governance.
Multimodal and generative vision: CLIP/BLIP/ALBEF, Vision Transformers, diffusion models (inpainting, super-resolution), visual RAG.
Performance tuning: ONNX/TensorRT, quantization, distillation, GPU scheduling; Numba/Cython.
Security and compliance basics for ML systems, PII handling, secrets management.
A/B testing, causal inference, and product analytics.
WE OFFER YOU
Flexible working time - you can agree on it within the team
Necessary tools and equipment
Communication in English - only foreign customers, and international teams
Simple structure and 'open door' way of communication
Full-time English teachers
Medical insurance for employees
HiQo University- internal education and training programs
HIQO COINS - We have a system of rewarding employees for extracurricular activities
#J-18808-Ljbffr