Applied AI Engineer

About Block Labs

Block Labs is a leading technology studio at the forefront of Web3, Artificial Intelligence, and iGaming innovation. We focus on creating robust, high-scale platforms that enable the future of digital products. Our team consists of senior engineers and strategists committed to top-tier system architecture, whether building autonomous AI agents, decentralized finance infrastructure, or high-frequency gaming systems. We prioritize long-term quality alongside rapid development.

Role Overview

As part of our shift towards integrating governed AI agents directly interacting with customers and internal teams, this role involves enhancing our data and intelligence platform. Our decision engine governs over 100 decisions across multiple modules, including risk management, user engagement, payments, treasury alerts, and responsible gaming, mixing human oversight with AI-driven optimization. You will engineer production-grade AI agents that operate autonomously but within secure, auditable boundaries, controlling real-money operations.

Key Responsibilities

• Develop Slack-integrated SQL BI analyst agents that reliably answer business queries using governed SQL over analytical databases with validation and traceability.
• Extend agent capabilities to customer-facing domains, implementing intent routing, retrieval-augmented generation (RAG) with versioned knowledge bases, multi-turn conversations, localized communication styles, and integration with customer support and CRM systems.
• Create risk-aware tool layers that validate information before action, support multi-step confirmation workflows, and escalate interventions appropriately while safeguarding against adversarial inputs and exploits.
• Build autonomous agents progressing through autonomy levels using frameworks such as LangGraph or Anthropic Agent SDK, including audit logging, feedback loops, and regression testing to ensure robust operation.
• Design, deploy, and maintain machine learning models addressing churn prediction, player scoring, fraud detection, anomaly identification, and other engagements, ensuring production readiness with monitoring and retraining pipelines.
• Lead data science initiatives involving multi-factor risk scoring, evidence-based policy codification with auditable, configurable rules, simulation and backtesting, and in-depth analysis to inform both agents and executive decision-making.
• Develop dashboards and user interfaces for supervisors to review agent actions, approve critical decisions, and monitor performance metrics, incorporating AI-assisted anomaly detection and user-friendly reporting views.

Preferred Qualifications

• Minimum of 4 years in software engineering, data science, or machine learning roles, with at least 1 year experience developing production AI agents leveraging large language models (LLMs) and orchestration frameworks.
• Experience delivering production retrieval-augmented generation systems, with a deep understanding of grounding techniques, hallucination mitigation, and refusal handling.
• Strong security mindset for customer-facing AI, including defenses against prompt injection, tool misuse, and data leakage.
• Ownership of ML model lifecycles from engineering through serving and monitoring; experience with fraud or risk detection is highly valued.
• Expertise in experimental design, uplift measurement, threshold calibration, and evaluation frameworks for non-deterministic AI systems.
• Proficiency in Python for production code, TypeScript for front-end review systems, and strong SQL skills—preferably on columnar databases such as ClickHouse.
• Capability to independently deliver stakeholder-facing tools, dashboards, and approval interfaces using modern frameworks or low-code tools like Streamlit.
• Experience designing systems that maintain a clean separation between probabilistic model outputs and deterministic logic execution, with observability and tracing tools such as Langfuse or LangSmith.

Desirable Extras

• Background in iGaming or transactional environments emphasizing security, auditability, and robust operational controls.
• Familiarity integrating with helpdesk or customer service platforms (e.g. Intercom, Zendesk) using APIs, webhooks, or automation workflows.
• Knowledge of blockchain or crypto transaction processes including on-chain analytics and stablecoin settlements.
• Experience with advanced optimization methods such as constrained optimization, bandit algorithms, or reinforcement learning applied to business rules.
• Hands-on with rule engines, decision-management software, or Slack app development.
• Competence in event-driven architectures and streaming platforms like Kafka or MSK, including fault-tolerant processing.

Working Environment

• Fully remote role with asynchronous communication; overlap with European time zones preferred.
• Small, autonomous Intelligence team within the broader Data function, working collaboratively across AI, BI, Infrastructure, and Customer Support teams.
• Design decisions are openly reviewed and debated; domain ownership is encouraged.
• Agent autonomy is progressively earned through rigorous evidence and gating mechanisms.
• Systems are built for scalable multi-tenant usage from the outset.

Culture

Our culture is mature, mission-focused, and low ego. We prioritize clarity, effective outcomes, and steady pace without chaos. We seek experts who want to collaborate with other top professionals in a respectful, high-standard environment.