Method is a global design and engineering consultancy founded in 1999. We believe that innovation should be meaningful, beautiful and human. We craft practical, powerful digital experiences that improve lives and transform businesses. Our teams [based in New York, Charlotte, Atlanta, London, Bengaluru, and remote] work with a wide range of organizations in many industries, including Healthcare, Financial Services, Retail, Automotive, Aviation, and Professional Services.
Method is part of GlobalLogic, a digital product engineering company. GlobalLogic integrates experience design and complex engineering to help our clients imagine what’s possible and accelerate their transition into tomorrow’s digital businesses. GlobalLogic is a Hitachi Group Company.
We are seeking a highly experienced and versatile Staff ML Engineer to join our dynamic Data & AI team. In this critical role, you will design, develop, and deploy AI applications incorporating sophisticated AI agents capable of complex reasoning, planning, and autonomous action. We’re looking for a candidate with hands-on expertise in Python—particularly using LangChain,LlamaIndex, HayStack, AutoGPT—who can architect agentic workflows and integrate complex ETL pipelines. You should be comfortable designing and deploying web app POCs on cloud and orchestrating tooling, APIs and third-party systems to build robust data-driven applications. You will tackle ambiguous problems, provide technical leadership across front-end, back-end, and AI components, and help shape the technical direction for our agentic AI initiatives. This is a unique opportunity to apply your full-stack expertise to the forefront of AI development.
Travel for team and client meetings is required, typically up to 15%. This is a UK-based position.
Responsibilities:
- Develop and manage AI-based data pipelines, including data curation, generation, and labeling for training and adapting AI models.
- Work with foundation models and unstructured data, performing tasks such as data deduplication, vectorization, tokenization, context retrieval, and quality control.
- Optimize AI model inference for speed and cost-effectiveness.
- Participate in the development and evaluation of AI applications, mitigating risks and identifying opportunities for performance enhancements.
- Evaluate AI models, particularly open-ended, non-deterministic models, ensuring appropriate tone and content.
- Apply prompt engineering techniques to guide AI models through input, without adjusting model weights.
- Construct effective instruction and context for AI models to perform tasks, enhancing model performance.
- Develop AI-powered interfaces such as browser extensions, chatbots, plugins/addons, web/desktop/mobile applications.
Skills Required:
- 5+ years experience in data science. Your story/portfolio reflects a solid foundation in using data and applying statistical/ML models to solve business problems. You demonstrate a good understanding of commercial data reality across many industries.
- Undergraduate and/or Masters Degree in Data Science or equivalent professional experience
- Experience in data engineering, including data curation, generation, and labeling (Pandas, NumPy, Sklearn). DataBase interfacing: SQLAlchemy, Alembic. Vector DBs: one of ChromaDB, Weaviate, Pinecone.
- Familiarity with foundation models and handling unstructured data. Experience with at least two of LangChain, LangSmith, llamaindex, OpenAI apis, Ollama, HuggingFace Transformers, CrewAI.
- Understanding of data preprocessing techniques like deduplication, tokenization, context retrieval, and quality control.
- Proven ability to optimize model inference for speed and cost-effectiveness, MLOps (MLflow, Tensorboard, Neptune, Weights and Biases)
- Experience in evaluating AI applications, identifying risks, and uncovering opportunities for enhancement.
- Skill in evaluating open-ended, non-deterministic AI models and ensuring appropriate tone and content.
- Proficiency in prompt engineering and context construction for guiding AI model behavior.
- Experience developing AI-powered interfaces such as browser extensions, chatbots, and web applications (with one of Flask, Jinja, FastAPI, Streamlit).
- Knowledge of various AI-related technologies and tools for orchestrating Agents, augmenting generative models, monitoring performance, and creating output guardrails.
Why Method?
We look for individuals who are smart, kind and brave. Curious people with a natural ability to think on their feet, learn fast, and develop points-of-view for a constantly changing world find Method an exciting place to work. Our employees are excited to collaborate with dispersed and diverse teams that bring together the best in thinking and making. We champion the ability to listen, and believe that critique and dissonance lead to better outcomes. We believe everyone has the capacity to lead and look for proactive individuals who can take and give direction, lead by example, enjoy the making as much as they do the thinking, especially at senior and leadership levels.
We believe in work/life balance. Seriously. We offer a ton of competitive perks, including:
- Continuing education opportunities
- Flexible PTO and work-from-home policies
- Pension
- Health and dental benefits
- Company lunches, company outings, along with a lot of snacks
- Health and wellness programs
- Other location specific perks (just ask!)
Next Steps
If Method sounds like the place for you, please submit an application. Also, let us know if you have a presence online with a portfolio, GitHub, Dribbble or other platform.
By clicking the "Submit Application" button below, you give your express consent to process your data for this vacancy according to the Recruitment Privacy Notice.
By checking the box, you may also agree to allow GlobalLogic to retain your data for future opportunities for employment for up to 3 years.
You can withdraw your consent at any time by contacting [email protected].
You can find more information about how GlobalLogic processes your personal data and what your rights are in the Recruitment Privacy Notice.
