Make $25 - $35 Per Hour Doing Simple Jobs From Home.
What is your current level of experience in freelance work?
Deploy ML Model on AWS with Time Series CSV Training & Inference Pipeline
Project Overview: I’m looking for an experienced AWS + Machine Learning + DevOps engineer to build a fully automated ML pipeline on AWS. The project involves loading a machine learning model (of my choice), training it on time series CSV data, running inference via API, and automating infrastructure provisioning using Terraform. Project Goals: Infrastructure-as-Code (IaC): Create a Terraform script to deploy all resources needed for the ML pipeline. Support terraform apply to spin up the full pipeline and terraform destroy to tear it down. Model Hosting: Set up an environment using AWS (SageMaker, S3, Lambda, EC2, or appropriate combo). Load a user-specified ML model (PyTorch, TensorFlow, etc.). Data Pipeline: Accept time series data via CSV uploads to S3. Trigger training pipeline automatically or via API/CLI. Model Training: Set up a training job (SageMaker preferred) using the uploaded CSV files. Enable model re-training and versioning. Inference Endpoint: Provide a secure endpoint (API Gateway + Lambda or SageMaker endpoint) for inference on new data inputs. Return predictions in a clean format (e.g., JSON). Deliverables: Terraform scripts that can: Provision all AWS resources Configure access roles and policies Destroy the environment cleanly End-to-end ML pipeline: Model deployment Training trigger with CSV upload Inference via REST API Documentation covering: How to use Terraform scripts Model training workflow Inference usage (with example request/response) Preferred Skills: AWS (SageMaker, Lambda, S3, API Gateway, IAM, EC2) Terraform (modules, outputs, variables, state handling) ML Frameworks: PyTorch / TensorFlow Python for data handling & model logic MLOps & secure cloud practices Time series ML experience is a bonus What I Will Provide: Model architecture or saved model file Sample CSV training & inference data Example expected outputs Optional (Bonus Points): Lightweight frontend to upload CSVs and view predictions CI/CD workflow using GitHub Actions or similar Logging & monitoring via CloudWatch