Software Architect

What is this project about?

An AI-powered analytics platform built for the defense domain. It ingests and analyzes data from multiple sources, automatically surfacing key findings and generating actionable insights and recommendations. By automating heavy manual analysis, the platform helps organizations make faster, better-informed decisions, reduce manual effort, and improve overall operational performance.

What is the team size and structure?

Project Manager, Backend and ML Engineers, UI/UX Designer, DevOps.

Hard skills

● 7+ years of experience in Software Engineering;
● Strong experience designing complex distributed systems and service-oriented architectures;
● Hands-on experience with AI/ML system architecture, including LLM-based systems, inference layers, vector search, and data pipelines;
● Solid understanding of LLM reasoning, tool calling, agent orchestration, and human-in-the-loop workflows;
● Ability to decompose complex platforms into services, modules, and integration layers, and to define clear boundaries between LLM, ML, deterministic logic, and human validation;
● Strong understanding of data architecture: data lakehouse, vector databases, object storage, and federated queries;
● Experience designing for scalability, latency, throughput, and fault tolerance;
● Experience designing secure systems for sensitive or critical data, with awareness of AI security risks (prompt injection, unsafe tool execution, data leakage, runtime protection);
● Strong documentation skills: architecture diagrams, ADRs, technical specifications, and integration flows;
● Ability to work with incomplete information, technical uncertainty, and evolving requirements;
● High level of ownership, technical leadership, and proactive decision-making;
● Upper-Intermediate level of English.

Responsibilities

● Design end-to-end system architecture for the AI-based data analysis platform;
● Decompose the system into services: agent runtime, orchestration, inference layer, data layer, security layer, and integrations;
● Define boundaries between LLM reasoning, ML processing, deterministic rules, and human decision-making;
● Design architecture for AI recommendations, next-best-action logic, semantic search, and data analysis workflows;
● Design a scalable inference architecture and model-serving approach;
● Define latency, throughput, fault-tolerance, and scaling strategy from single-node to cluster deployment;
● Design data architecture for ingestion, storage, cataloging, vector search, querying, and access control;
● Define the orchestration approach for long-running AI/agent workflows;
● Establish security, auditability, runtime protection, and data leakage prevention principles;
● Collaborate with engineering, DevOps, ML/data engineers, product, and business stakeholders;
● Support technical discovery, feasibility analysis, POC creation, concept validation, estimation, and roadmap planning;
● Identify architectural risks, dependencies, constraints, and mitigation options;
● Prepare and maintain architecture documentation, diagrams, ADRs, and technical recommendations;
● Guide engineering teams on implementation approach, scalability, performance, and architectural standards.

Nice to have

● Experience in the Defense/MilTech and AI domains;
● Experience with on-premise/hybrid infrastructure.

How many stages of the interview are there?
● Recruiter interview – up to 30 minutes;
● Technical interview – up to 1.5 hours;
● Polygraph screening.