AI engineer python

грейд
senior
формат
гибрид Лондон
язык
Английский B1
опубликована
сегодня

описание

We’re seeking someone to join our team as an AI Engineer using Python to develop the firmwide Enterprise AI Platform, which aligns with the firm's Technology principles and drives efficiency and consistency, controls, security and strong governance and promotes innovation, enabling teams to build applications that leverage AI capabilities and accelerate the adoption of AI across our businesses.

задачи

  • Deploy to partner technology teams. Join a partner team for a short, bounded engagement aimed at one problem. Run discovery with the people who own the process, and establish what is genuinely manual, brittle, or held back by policy rather than by technology. Agree scope, a measure of success and an exit test before you start.
  • Assess before you build. Decide whether the problem suits an agent and recommend a conventional application or scripted automation where that is the better answer.
  • Build agentic systems in development. Use the firm's agent orchestration, evaluation and model-access platforms together with open agent frameworks and protocols: task decomposition, tool and context design, orchestration across sub-agents, and state and session management for long-running work.
  • Choose designs proportionate to the problem that the inheriting team can operate, explain and extend.
  • Establish the evaluation suite. Work with the team's subject-matter experts to build stratified test sets, define numerically what acceptable performance means, grade the full sequence of steps an agent takes, and improve against the suite measurably.
  • Build governance in from the outset. Place human approval according to the consequence and reversibility of an action, add policy checks that cannot be bypassed, and keep an audit trail an examiner can follow. Route policy decisions to their owners with the options laid out.
  • Hand over so the system outlives the engagement. Pair with the team's engineers from day one, advise them through testing and acceptance, and leave named owners, runnable documentation, a published case study, and a team that can deploy, evaluate and extend the system alone.
  • Report back to the central team. Tell the product teams where our products fall short today: gaps, friction, missing capabilities and the workarounds teams adopted, backed by evidence from the engagement, so they can plan their roadmaps.
  • Grow the pattern library. Contribute common patterns, reference implementations and reusable components.

требования

  • Bachelor's or a master's degree in computer science or related field, or equivalent job experience
  • 4-6 years of experience in software engineering, design, and development
  • Experience building AI or agentic applications that were deployed to real users
  • Strong hands-on proficiency in Python, including writing well-structured, reusable modules and clear interfaces for other engineers
  • Working knowledge of agent engineering: tool and function calling, Model Context Protocol, retrieval and long-context tradeoffs, prompt and context management, and common orchestration patterns
  • Practical experience evaluating AI systems: automated assertions, model-graded scoring against explicit criteria, test-set construction and regression checks
  • Familiarity with running AI systems in production: observability and tracing, common failure modes, reproducibility, cost awareness, and upgrading model and dependency versions
  • Working knowledge of SQL and at least one non-relational or large-scale data store, and awareness of data governance, data privacy and security requirements
  • Demonstrated experience in DevOps, understanding of CI/CD, and hands-on experience managing code in repositories such as GitHub and Bitbucket
  • Ability to join an unfamiliar codebase and get a change through it
  • Ability to frame a problem before designing a solution, to say what is known and unknown before committing to a claim, and to tell whether a solution is a prototype or built for production
  • Ability to articulate technical concepts effectively to diverse audiences, including business stakeholders and control functions
  • Strong interest in teaching and in helping other development teams adopt AI, with some experience presenting, mentoring or running a technical session
  • Demonstrated ability to work effectively and collaboratively in a global organization, across time zones, and across organizations
  • Будет плюсом: Experience delivering inside a regulated business process such as client onboarding, trade control, credit, or audit; Experience working with a control function (model risk, compliance, or internal audit) from either side of the review; Experience designing human-in-the-loop controls, approval gates, or audit evidence for an automated decision process; Experience working with cloud providers, preferably Azure, but also AWS and GCP, and with their managed AI development services; Experience in an embedded or client-facing technical role, or working alongside external vendor or consultant engineers; Contributions to shared frameworks, libraries or open source that other engineers adopted; Understanding of Model Ops / ML Ops / LLM Ops

условия

  • Our teams are relentless collaborators and creative thinkers, fueled by their diverse backgrounds and experiences. We are proud to support our employees and their families at every point along their work-life journey, offering some of the most attractive and comprehensive employee benefits and perks in the industry.
  • There’s also ample opportunity to move about the business for those who show passion and grit in their work.
  • Morgan Stanley empowers employees to have greater freedom of choice through flexible working arrangements.

навыки

компания

похожие вакансии

статьи для разработчиков