Tiquo is building a connected operating platform for hospitality and service businesses. We bring bookings, point of sale, payments, memberships, customer relationships, inventory and reporting into one system, helping hotels, restaurants, spas, members’ clubs and venues replace disconnected tools with simpler operations and a consistent guest experience.
About the role
We’re looking for an ML Platform Engineer to design and build the data, inference and evaluation infrastructure behind Tiquo’s AI capabilities. This is a full-time, hybrid role based in the London Area, United Kingdom.
This is a hands‑on engineering role spanning distributed systems, data engineering, MLOps and production AI. You’ll work closely with our software engineers to turn transactional and operational data into secure, testable AI services, supporting natural‑language analytics, forecasting and workflow automation. The emphasis is on production systems, not standalone notebooks or model research.
You’ll own technical decisions across ingestion, dataset and feature versioning, model serving, retrieval, evaluation and observability, with a focus on correctness, tenant isolation, latency and cost.
What you’ll do
- Design batch and streaming pipelines for bookings, orders, payments, customer activity and inventory, with schema validation, idempotent processing, deduplication, replay and backfills.
- Build reproducible dataset and feature pipelines with lineage, versioning and point‑in‑time correctness; prevent data leakage and training‑serving skew.
- Implement model and AI‑service deployment pipelines covering artifact management, configuration, automated tests, staged releases and rollback.
- Build inference APIs and asynchronous workers with batching, caching, timeouts, retries, rate limits and graceful degradation; benchmark throughput, tail latency and cost per request.
- Develop retrieval and tool‑calling infrastructure for language‑model applications, including embedding pipelines, index updates, structured outputs and permission‑aware access to operational data.
- Build safe natural‑language analytics workflows with schema grounding, query validation, read‑only execution, tenant‑scoped access and auditable results.
- Create offline evaluation datasets and regression suites for model quality, retrieval relevance, factual grounding and tool execution; gate releases on measurable quality and performance criteria.
- Instrument pipelines and inference services with metrics, logs and distributed traces; monitor data quality, drift, failure rates, latency and resource utilisation.
- Enforce tenant isolation and role‑based access throughout retrieval, feature generation and inference; apply data minimisation, secrets management and safeguards against prompt injection and unauthorised actions.
- Diagnose production failures, profile bottlenecks, plan capacity and document architecture, deployment procedures and incident runbooks.
What you’ll bring
- Strong Python and SQL skills, with experience building tested, maintainable production services and data‑processing code.
- A solid understanding of distributed systems, including delivery semantics, consistency, concurrency, queue‑based processing and failure recovery.
- Hands‑on experience with relational databases, data modelling, indexing, query optimisation and analytical workloads.
- Experience with pipeline orchestration, dataset or feature versioning, model registries and reproducible ML workflows.
- Experience deploying and operating inference services, including API design, containerisation, CI/CD, Infrastructure as Code and cloud infrastructure.
- Practical knowledge of machine learning evaluation, temporal validation, data leakage, drift and the trade‑offs between model quality, latency and cost.
- Experience integrating language models into production applications, including retrieval‑augmented generation, tool calling, structured outputs and systematic evaluation.
- A security‑minded approach to multi‑tenant systems, access controls and handling sensitive customer and operational data.
- Confidence debugging across application code, pipelines, databases and infrastructure, and communicating technical decisions clearly.
Experience with technologies such as PyTorch, MLflow, Airflow or Dagster, Kafka, Kubernetes, Terraform, vector databases or OpenTelemetry would be useful. Depth of engineering understanding matters more than experience with every tool. Experience with transactional SaaS, hospitality systems or payment integrations is a plus. We value practical engineering judgement and relevant experience, whether gained through formal education or hands‑on work.
You’ll help shape the architecture and engineering practices behind AI in a platform that runs real‑world hospitality operations. You’ll work on technically demanding problems at the intersection of connected operational data, distributed systems and production AI, building capabilities that staff and businesses can trust.
#J-18808-Ljbffr
