Senior AI Engineer
Betfred is seeking a Senior AI Engineer to join its technology department. This is a hands-on engineering role focused on integrating LLMs and coding agents into existing software development workflows to improve efficiency and delivery speed. The role involves identifying where AI provides genuine value, building necessary tooling, and measuring the impact of these implementations within real codebases.
The successful candidate will work closely with engineering teams to design and run pilots, build integrations, and establish evaluation frameworks. Betfred provides the backing of the Engineering Director and the CIO, along with a budget for tools and the freedom to conduct experiments. The role requires a pragmatic approach, where the engineer is expected to make evidence-based decisions, report results transparently, and discontinue initiatives that do not deliver measurable benefits.
- Map how work flows through the engineering lifecycle, from idea to production, and find where effort, delay and rework actually concentrate.
- Establish baselines before changing anything, working with engineering, delivery and product teams: cycle time, review time, deployment frequency, change failure rate, time to restore, and developer feedback.
- Review the AI tools and initiatives already in use or under way across engineering, assess each against evidence of value and cost, and drive forward those that deliver.
- Close those that do not, or that can be solved better by other means such as conventional automation, a process change or an existing tool.
- Design and run pilots, with comparison groups where practical, covering AI coding assistants, coding agents, automated code review, test generation, bug triage and fixing, migrations, documentation and onboarding.
- Build and maintain the integrations that make AI useful in the environment: repository context and instruction files, CI/CD hooks, review assistants, agent workflows that open draft pull requests, and access to issue trackers, logs and documentation through APIs or MCP.
- Own the evaluation harness for everything built: test sets, regression checks, tracing and cost monitoring, so we know how often agents are right, what they cost, where they fail, and whether fixed bugs stay fixed.
- Define how engineers use AI safely and well: pull request size limits, test requirements, accountability for AI-assisted code, and what code and data may go to which tools.
- Work with Security, Legal and Data Protection on vendor, IP, privacy and code-security controls.
- Evaluate hosted and self-hosted model options against requirements for code and data security, and recommend what should be run where.
- Coach engineers and win adoption, especially from sceptics, while protecting the development of junior engineers.
- Choose the right tool for each problem: an LLM, an agent, conventional automation (linters, codemods, better CI) or no change at all, and decide when to build and when to buy.
- Build the business case: quantify the costs and benefits of what is built, show the return on investment to engineering and executive leadership, and recommend what to scale, change or stop.
- Report results openly, including costs and failures. Honest negative findings are valued.
- Strong professional software engineering experience, including building and operating production software (required)
- Ability to read and critically assess code outside of one's primary language (required)
- Excellent programming and software design skills with a strong grasp of testing, CI/CD, code review and modern development practice (required)
- Hands-on experience building applications with LLM APIs, including tool calling and agentic workflows (required)
- Practical understanding of how LLMs and coding agents fail: plausible but wrong code, fixes that mask symptoms, tests gamed to pass, context limits, and cost growth (required)
- Day-to-day experience using AI coding tools and agents on real codebases, with an evidence-based view of where they help and where they hurt (required)
- Experience integrating with developer systems such as Git platforms, CI/CD, issue trackers and observability tooling (required)
- Experience designing evaluations for AI outputs, or comparable experimental rigour (required)
- Comfort preparing code, documents and tickets for retrieval using Python and SQL, and working with a data team on anything larger (required)
- Credibility with engineers, and the ability to influence without line authority (required)
- A pragmatic, experimental mindset, including willingness to stop your own ideas when the evidence says so (required)
- Experience with Python or a comparable language for tooling and AI services, and comfort working across languages and codebases (required)
- Experience with LLM APIs and open-weight models (provider-hosted or self-hosted), tool/function calling, Model Context Protocol, agent frameworks, and AI coding tools and agents (required)
- Experience with developer productivity or developer experience engineering, DORA-style measurement, platform engineering, RAG and vector search, deploying and evaluating self-hosted open-weight models, or experience in a regulated or customer-facing environment such as gambling or financial services (nice-to-have)
- Experience running open-weight models yourself, fine-tuning, or working with embeddings (plus)
Betfred is a betting and gaming company that has operated for over 50 years, founded by Fred and Peter Done. It runs more than 1,300 betting shops and employs around 9,500 people, combining a large retail estate with online sports betting and gaming. In 2019 it entered the US sports betting market as Betfred Sportsbook, now operating in several states, and has established a presence in South Africa, where it holds a majority stake in the online betting company LottoStar. Headquartered in Warrington, England, Betfred emphasises customer service and responsible gambling across its operations.