Senior MLOps Engineer

Description :

Senior MLOps Engineer.

Experience : 6+ Years.

Location : Remote.

Role : Contract.

We are looking for an MLOps Engineer with hands-on experience in deploying and scaling ML inference systems in production. You will set up and deploy systems at scale to support a pipeline that transforms millions of photos, videos, and text documents into searchable embeddings using a combination of deep learning models (DistilBERT, SBERT, TransNetV2) and external multimodal APIs. The ideal candidate has experience optimizing inference at scale, orchestrating ML workloads, and working with both PyTorch and TensorFlow in a cloud environment.

What Youll Do :

- Prepare SageMaker pipelines for workflows coming in Q1 including reranking of search queries.

- Working with Dev and DevOps teams, set up end-to-end SageMaker pipelines for model tuning and monitoring in Dev, QA, and Prod environments.

- Work with data science team to roll out improvements.

- Deploy and monitor the GPU instance for a reranker to support hundreds of thousands of search queries per day.

- Verify and confirm optimal GPU instance type for the reranker.

- Working with Dev and DevOps teams, deploy reranker in QA and Prod.

- Set up monitoring for reranker.

- Set up the environment and instance(s) for TransnetV2 to perform video shot detection at scale.

- Containerize and autoscale TransnetV2 and manage massive I/O to avoid CPU bottlenecks.

- Ensure model inference, orchestration and monitoring (latency/failure rate for video types, drift detection).

- Establish a standard for ML Ops model deployments at AP, including provisioning appropriate

infrastructure and orchestration.

- Create deployment pipelines for each of the models in the current Prod stack and in active development.

- Working with Dev and DevOps teams, deploy models to Dev, QA and Prod.

Requirements :

- Experience with both computer vision and language models strongly preferred.

- The ability to select and transform features at scale for BERT-based models.

- Setting up monitoring and evaluation metrics in Sagemaker.

- Creating A/B tests for ML models in Sagemaker.

- Expertise in model deployment and orchestration.

Experience :

- Deploying transformer-based models with Tensorflow or PyTorch in production workflows.

- Resource selection, evaluation and optimization for production model deployments.

Familiarity and demonstrated experience with the following strongly preferred :

- Feedforwards Neural networks.

- Diffusion Models.

- Ranking Algorithms.

- Approximate Nearest Neighbors HNSW.

- Bayesian Methods.

- Regression Models.

- Decision Trees.

- Clustering (K-Means, DBScan).