We are an international group of companies operating at the intersection of fintech, blockchain, and iGaming. We develop our own technological products and solutions for global markets, including payment infrastructure, analytical systems, and gaming platforms.
The company actively invests in promising projects in the field of digital finance and crypto-technologies, as well as creates and scales products for the iGaming industry, providing them to international clients.
We have a strong technical team and a product approach, building our own solutions from scratch. Over 10 years in the market, we work globally and continue to grow actively.
Responsibilities:
- Develop and implement isolation and control mechanisms for agent systems (sandboxes for tool execution, strict isolation of environments, action limits, and implementation of a kill switch for emergency agent shutdown).
- Configure detailed audit logs of agent actions in production.
- Develop and configure a system for monitoring abnormal behavior and alerting on suspicious patterns.
- Conduct threat modeling for new agent scenarios at the design stage.
- Protect against LLM attacks: develop architectural solutions and filters to protect against prompt injection, data exfiltration, jailbreak. Implement strict validation of agent output validation.
- Organize and conduct attacks on internal agent systems to identify vulnerabilities before attackers do.
- Conduct security audits of agent pipelines from the LLM engineering team to production rollout.
- Control outgoing data: ensure control over what information goes to external LLM providers and compliance with data usage policies. Ensure inventory and control of all data flows.
- Configure IAM and secrets management: design and implement a role-based access model for the agents themselves. Determine which agent has access and rights to perform actions in which environment. Configure secure storage and rotation of secrets.
- Use source code scanners: select, pilot, install, administer, and integrate source code scanners (SAST/DAST) into CI/CD pipelines. Decipher scanning results for developers, assist in resolving code vulnerabilities.
Candidate Profile:
- Experience: 3–5 years in DevSecOps, AppSec, or Security Engineering.
- Proficient in Python, deep experience with Docker, Linux, understanding of CI/CD principles.
- Practical threat modeling experience, not just theoretical knowledge.
- Experience with secrets management systems (HashiCorp Vault or similar), implementing IAM models, configuring sandboxing and isolated environments.
- Understanding of how LLMs and agent systems work. Experience with frameworks like LangChain, LangGraph, MCP, understanding the principles of Claude Code.
- Willingness to not only review others' code but also to independently design and build secure agents.
Bonus points:
- Knowledge of AI security specifics: OWASP LLM Top 10, experience with AI Red Teaming, practical skills in protecting against prompt injection.
- Experience with APIs of leading LLM providers (Anthropic, OpenAI) or with open-source models.
- Understanding of anti-fraud logic and how agents can both break it (create new attack vectors) and strengthen it (create new defense rules).
We guarantee our future colleague:*
- Competitive salary: 250,000 — 400,000 ₽/month net.
- Official employment according to the Labor Code from day one.
- Continuous salary indexing.
- Corporate equipment, modern foreign software.
- Stunning office in Moscow City.
- Bureaucracy-free environment.
- Office-based work format 5/2 from 09:00 to 18:00 or from 10:00 to 19:00.
- Work in an ambitious team and rapid professional development.
