In the rapidly evolving field of artificial intelligence and machine learning (AI/ML), there’s a common perception that expertise in model training is the cornerstone of a successful career. While understanding how to develop and train models certainly holds value, it’s important to recognize that a wide array of essential roles contribute to the successful deployment and operationalization of AI/ML solutions.
I read many articles, books, and forums related to AI and Machine Learning. In most of my reading, the authors point people to linear algebra and calculus as the starting point for a career in Machine Learning. While this is true for those scientists and engineers who want to train models or even fine tune LLM’s, it is not true for many roles that are necessary to productionize and operationalize machine learning applications in the entperprise.
From solutions architects and backend developers to product managers and support engineers, each role plays a critical part in ensuring that AI technologies are effectively integrated into business processes, delivering real value without requiring specialized knowledge of model training. This article explores these diverse roles, highlighting how collaborative efforts and varied skill sets are fundamental to transforming AI/ML projects from concepts into impactful solutions.
Consider these roles, for example:
Solutions Architect
A Solutions Architect would design a scalable, secure, and efficient architecture that aligns with both technical and business requirements. They would ensure seamless integration of ML models or Gen AI applications with existing systems, manage infrastructure for performance and scalability, and apply security and compliance standards to protect data. By implementing MLOps practices, they would streamline model deployment, monitoring, and retraining to maintain reliability. Additionally, they work closely with data scientists, MLOps engineers, and stakeholders to balance technical feasibility with cost management, making the AI solution sustainable and aligned with business objectives. In all of these functions, they would not need to understand the linear algebra or calculus that is used to build the model.
Backend Developer
A backend developer in a machine learning or generative AI project would focus on creating the infrastructure and APIs necessary for serving model predictions and integrating the model into applications. They would build and maintain endpoints that allow other services or users to interact with the model, handling data input and output securely and efficiently. They would optimize performance for low-latency responses, manage data transformations for model compatibility, and implement error handling to ensure a robust user experience. Additionally, they would work on logging and monitoring to capture metrics on model usage and performance, enabling continuous improvement and scalability of the deployment environment. While all of these functions are important in productionizing ML and Gen AI, it is not necessary for a backend engineer to understand how to train a machine learning model.
Frontend Developer
A frontend developer in a machine learning or Generative AI project would be responsible for designing and implementing the user interface through which users interact with the model’s predictions or outputs. They would build intuitive, responsive, and accessible UIs that enable users to input data, view results, and receive feedback seamlessly. This includes implementing API calls to fetch model predictions and displaying them effectively, whether as text, images, or interactive visuals. Frontend developers would also focus on enhancing the user experience by handling loading states, errors, and user interactions smoothly, ensuring the AI-driven features integrate naturally into the application’s overall design and functionality. These functions would not require skills in training a machine learning model, but they are important for implementing the project.
Infrastructure Engineer
An infrastructure engineer would be responsible for setting up and managing the underlying infrastructure to support the model’s deployment and performance. They would handle provisioning and configuring servers, storage, and networking resources to ensure the model can deliver low-latency, high-availability predictions. Infrastructure engineers would focus on scaling, load balancing, and reliability, often using cloud services or container orchestration tools to automate and optimize resources as demand fluctuates. Additionally, they would implement monitoring, logging, and alerting to maintain visibility into system health, helping to ensure a stable and cost-effective environment for delivering AI-driven services to users. These skills are certainly important in an enterprise implementation of AI/ML, but the infrastructure engineer would not need to understand the math that forms the foundation of machine learning.
Support Engineer
A support engineer would provide frontline assistance to ensure the solution runs smoothly for end-users and internal teams. They would address user issues related to model functionality, troubleshoot performance or access issues, and escalate technical problems to developers or infrastructure teams when needed. Support engineers would gather and analyze user feedback on the model’s performance or integration issues, contributing to improvements by sharing insights with development and product teams. They may also create documentation, FAQs, and other resources to help users understand and maximize the AI solution’s capabilities while keeping the support process efficient and responsive. All of these functions are important for operationalizing a machine learning project but do not require an understanding of how to train an ML model.
Sales Engineer / Consultant
A sales engineer or consultant would bridge technical capabilities with client needs, helping to communicate the value of the solution in business terms. They would work closely with clients to understand their specific requirements, tailoring demonstrations or proofs of concept to showcase how the AI solution can address particular pain points. Sales engineers would help clients understand the deployment process, integration options, and ongoing support, providing insights into how the technology can be implemented effectively. They would also serve as liaisons between clients and the technical team, ensuring the solution is aligned with both business goals and technical feasibility, and fostering client confidence in the AI product’s impact and return on investment. While sales engineers and consultants are necessary for a successful AI/ML project, these people do not need to understand how to train machine learning models.
Product Manager
A product manager would oversee the project’s alignment with business goals and user needs, guiding the product’s development and deployment to deliver maximum value. They work with cross-functional teams—including engineering, design, and sales—to define requirements, prioritize features, and ensure the solution is both technically feasible and market-ready. Product managers would focus on understanding user feedback and setting key performance indicators (KPIs) to measure success, balancing innovation with usability and reliability. They would also help shape the go-to-market strategy and communicate the product’s benefits, ensuring that the AI solution is effectively positioned to meet client and user expectations and drive business impact. This is definitely an important function in an enterprise machine learning or Gen AI project, but it doesn’t not require knowledge of training machine learning models.
Conclusion
The AI/ML job landscape is changing almost as fast as the technology. In the near future, we will see more companies hiring roles that require an understanding of how to USE machine learning and generative AI, but without requiring an understanding of the science behind the models themselves.
