Solutions Architect - Machine Learning

The Role

We are seeking a Solutions Architect to join our Engineering Team. This role combines deep hands-on engineering capability in Python, machine learning frameworks, and data pipelines with technical leadership in designing, developing, deploying, and optimizing production-grade machine learning models and solutions.

The ideal candidate is a senior technical leader who can define end-to-end ML solution architecture, lead model development, evaluation, deployment, monitoring, and production support, and ensure adherence to ML engineering and MLOps best practices while delivering reliable, scalable, and maintainable solutions in a fast-moving environment.

This position also requires strong collaboration and leadership skills. The Solutions Architect must work effectively with data scientists, engineers, product stakeholders, clients, and delivery teams, while providing technical guidance, mentoring others, and driving high engineering quality. The ideal candidate should be proactive, pragmatic, and able to balance hands-on implementation with strategic technical decision-making.

Key Responsibilities

Solution Architecture: Lead the architecture of machine learning solutions end-to-end, from data ingestion and feature engineering to model serving and monitoring, ensuring scalability, security, and maintainability

Model Development: Design, develop, and optimize machine learning models using Python and ML frameworks, while leading evaluation and adoption of new tools, algorithms, and methodologies to improve company standards

Data Pipelines: Design and oversee reliable batch and streaming data pipelines and feature stores that feed training and inference, ensuring data quality, lineage, and reproducibility

Model Evaluation: Define evaluation strategies, metrics, validation approaches, and experiment tracking to ensure models meet accuracy, fairness, and business performance targets before release

Deployment & MLOps: Lead model packaging, deployment, and CI/CD/CT (continuous training) pipelines, including model registry, versioning, and automated promotion across environments

Monitoring & Production Support: Own production health of ML systems by monitoring model performance, data and concept drift, latency, and cost; lead incident triage, root-cause analysis, and retraining decisions

Best Practices: Establish and ensure adherence to ML engineering best practices, including code quality, testing, reproducibility, documentation, and responsible-AI and data-privacy requirements

Cross Functional Collaboration: Partner with business ops, stakeholders, and clients to translate business problems into ML solutions and act as the intermediary between business operations, data science, and engineering

Team Development: Provide technical guidance and mentor engineers and data scientists at all levels while also leading training sessions, design and code reviews, providing constructive feedback, and aligning technical standards across the team

Soft Skills

Business Acumen: Connect ML architecture and modeling decisions to business outcomes, anticipating impacts on cost, risk, and value, and providing decisions to maximize long-term value.

Accountability: Accountable for the technical and delivery success of ML solutions or projects, taking ownership of outcomes across teams and addressing issues proactively rather than reactively.

Communication: Communicate complex ML concepts, model behavior, and trade-offs clearly to both technical and non-technical audiences, aligning stakeholders, and enabling confident decision making.

Judgement: Demonstrate judgment by making high-impact decisions, balancing experimentation and short-term delivery with long-term sustainability, and escalating risks early.

Collaboration: Drive alignment across multiple teams and disciplines by acting as a unifying technical leader, resolving cross-team friction.

Curiosity: Maintain curiosity about the evolving ML landscape and emerging technologies, using that understanding to anticipate challenges, guide innovation, and continuously improve technical and delivery practices.

Required Qualifications

Experience: Over 6+ years of software or data engineering experience, including 3+ years in an architect or technical lead role and 4+ years delivering machine learning solutions to production, in the following:

Tech Stack

Languages & Core ML

Python (primary); SQL; Scala or Java a plus

NumPy, pandas, Polars

scikit-learn, XGBoost, LightGBM, CatBoost

Deep Learning Frameworks

PyTorch, TensorFlow / Keras

Hugging Face Transformers

Model optimization: ONNX, quantization, distillation

Data Engineering & Pipelines

Apache Spark / PySpark, Databricks

Orchestration: Apache Airflow, Prefect, Dagster, or Azure Data Factory

Streaming: Kafka, Azure Event Hubs, or AWS Kinesis

Data warehouses and lakehouses: Snowflake, Delta Lake, BigQuery

Data validation: Great Expectations, Pandera

MLOps & Model Lifecycle

Experiment tracking and model registry: MLflow, Weights & Biases

ML platforms: Azure Machine Learning, AWS SageMaker, Google Vertex AI

Feature stores: Feast, Databricks Feature Store, or equivalents

Pipeline frameworks: Kubeflow, Azure ML Pipelines, SageMaker Pipelines

Model Serving & APIs

FastAPI, Flask, or equivalent API frameworks

Serving: BentoML, KServe, TorchServe, Triton, or managed endpoints

Batch and real-time inference patterns, RESTful API design

Cloud & Infrastructure

Azure, AWS, or GCP cloud-native services

Containers and orchestration: Docker, Kubernetes

Terraform or Bicep for Infrastructure as Code

Architecture & Patterns

ML reference architectures (training, inference, feedback loops)

Event-driven and microservices-based ML systems

Reproducibility patterns: data and model versioning (DVC, Delta Lake)

DevOps

CI/CD/CT pipelines (Azure DevOps, GitHub Actions)

Git branching and pull request workflows

Testing & Quality

pytest, unit and integration testing for data and ML code

Model validation, bias/fairness testing, and explainability (SHAP, LIME)

Monitoring & Operations

Model and drift monitoring: Evidently AI, WhyLabs, Arize, or platform-native tools

Observability: Prometheus, Grafana, Application Insights, OpenTelemetry

Must have:

Proven experience designing, developing, deploying, and optimizing machine learning models in production using Python and ML frameworks (scikit-learn, PyTorch, TensorFlow, XGBoost).

Strong experience building and operating data pipelines for training and inference, including Spark / Databricks and workflow orchestration (Airflow or equivalent).

Hands-on MLOps experience: experiment tracking, model registry, versioning, and CI/CD/CT pipelines using MLflow and a major ML platform (Azure ML, SageMaker, or Vertex AI).

Track record leading model evaluation, deployment, monitoring (performance and drift), and production support for business-critical ML systems.

Experience deploying models as scalable services (batch and real-time) using containers, Kubernetes, and API frameworks such as FastAPI.

Track record leading solution architecture for enterprise-scale systems, including non-functional requirements, trade-off analysis, and architecture governance.

Demonstrated ability to define and enforce ML engineering best practices and to provide technical guidance and mentor engineering teams.

Preferred Qualifications

Experience with deep learning for NLP or computer vision, and with Hugging Face Transformers.

Experience with feature stores and real-time/streaming ML architectures (Kafka, Event Hubs).

Exposure to LLMs and generative AI, including fine-tuning, RAG, or LLMOps practices.

Familiarity with AI governance and risk frameworks (NIST AI RMF, ISO/IEC 42001), model explainability, and data privacy regulations.

Experience with Snowflake or lakehouse architectures (Delta Lake, Databricks Unity Catalog).

Consulting or client-facing delivery experience, including discovery and whiteboard sessions.

Cloud or ML certifications (e.g., Azure Data Scientist Associate, AWS Machine Learning Specialty / ML Engineer Associate, Google Professional ML Engineer).