ML Ops

We deliver value-driven solutions powered by a unique combination of our cutting-edge machine learning and 3D graphics engineering expertise.

Technologies

Jetson Nano

MLflow

Prometheus

Custom exporter

Filebeat

PostgreSQL

Filebeat

Logstash

Elastic Search

BERT

TensorFlow Serving

Triton Server

Docker

MLflow

AirFlow

KubeFlow

Jupyter Notebook

MS Azure

AWS

Google Cloud

Vertex AI

DialogFlow

Airflow

MLflow

K8s cluster

Jetson Nano

MLflow

Prometheus

Custom exporter

Filebeat

PostgreSQL

Filebeat

Logstash

Elastic Search

BERT

TensorFlow Serving

Triton Server

Docker

MLflow

AirFlow

KubeFlow

Jupyter Notebook

MS Azure

AWS

Google Cloud

Vertex AI

DialogFlow

Airflow

MLflow

K8s cluster

case study

NVIDIA Jetson

Assembling, environment creation, and fine-tuning for NVIDIA Jetson for AI applications.

Jetson Nano
ML Ops
Automotive
Manufacturing

case study

Monitoring

Develop interactive dashboards where one can access the performance of ML models during training and production with ease.

MLflowPrometheusCustom exporterFilebeatPostgreSQLFilebeatLogstashElastic Search
ML Ops
IT Services and Technology
Banking (Fintech)
Manufacturing

case study

ML Ops

Survey of different ML infrastructure building approaches.

BERTTensorFlow ServingTriton ServerDockerMLflowAirFlowKubeFlowJupyter NotebookMS AzureAWSGoogle CloudVertex AIDialogFlow
ML Ops

case study

ML experiments with Airflow

Goal is to convert CV models to different formats (ONNX, quantized, TVM compiled) and test them systematically.

AirflowMLflowK8s cluster
ML Ops