Senior Generative AI Engineer

About the role

Ajackus is building an AI-powered concierge platform for private members clubs, and we are looking for a Senior AI/ML Engineer to own its intelligence layer.

Members message their club on WhatsApp, under the club's own name. An AI agent replies, makes bookings, takes payments, and introduces each member to the people they would benefit from knowing, with a clear reason behind every introduction. The platform is launching with private members clubs and will expand to sports clubs, alumni networks, hospitality groups, and online communities.

You will own the AI service and its data architecture end to end, as the senior engineer in a small, focused product team at Ajackus.

Key responsibilities

• Member intelligence and matching: Design the data architecture for people and relationships, and build reason-based matching (business, investment, advice, shared interests, friendship) that explains every introduction and improves from feedback, proven by evals.
• Agentic concierge: Build end-to-end WhatsApp workflows for bookings, payments, club knowledge, and escalation, with reliable rules for when the agent acts, asks for confirmation, or hands over to a person.
• User journeys: Map and build journeys for members, club teams, and internal users, each with a defined state, next step, and success measure.
• Multi-tenant architecture: Enable new clubs and new verticals to be set up through configuration alone, with no code changes.
• Security and trust: Enforce per-club data isolation, consent tracking, full erasure, a restricted store for sensitive data, audit logs for every agent action, and defences against prompt injection and data leakage, building towards SOC 2 and ISO 27001.
• Performance and cost: Measure and reduce latency and token cost per turn through model routing, caching, prompt optimisation, and batching, while protecting quality with golden datasets, CI regression gates, and drift monitoring.

Must-have requirements

• 4+ years building production software, with substantial hands-on time on LLM systems that take real actions
• Experience designing data architecture for people and relationships: profiles, graphs, and explainable matches
• Recommendation, ranking, or matching systems in production, including systems that learn from feedback
• Agentic tool use in production: multi-step workflows with confirmation, escalation, and failure recovery
• RAG in production, with tenant-scoped retrieval and answers grounded in their sources
• Multi-tenant AI systems with per-tenant data isolation and configuration
• Security and privacy by design: access control, consent management, UK GDPR, data erasure, and audit trails
• A proven, measurable record of reducing LLM latency and token cost without losing quality
• Deep Python expertise and comfort owning an async production service
• Exceptional written English and strong attention to detail in user-facing copy
• Experience as the senior or sole owner of an AI service in a small team
• Willingness to work directly with the backend when the agent needs a new API contract

Tech stack

Python, FastAPI, LLM APIs (generation and embeddings), PostgreSQL with pgvector, Redis, AWS

About Ajackus

Ajackus is a technology consulting company with over 15 years of experience delivering software engineering, AI, cloud, and data solutions to startups and enterprises across the globe. We specialize in building high-performing engineering teams and solving complex technology challenges for our clients.

If you're passionate about software engineering, enjoy evaluating technical talent, and want to contribute to building world-class engineering teams, we'd love to hear from you.