Senior AI Engineer

CAST; Bengaluru; Hybrid; INR 250,000 - 450,000; Full time

Job summary

CAST R&D in Paris and Bengaluru seeks a Senior AI Engineer to lead generation AI capabilities on CAST's deterministic software-intelligence data, shaping architecture across services for developers, architects, and executives.

You will design scalable, fault-tolerant RAG pipelines, integrate LangChain tools, and guide multi-agent orchestration, balancing performance and cost. You will mentor engineers and help set responsible-AI standards while delivering measurable insights.

Qualifications

• 4+ years hands-on experience building and shipping ai/ml systems in production.
• Experience building llm applications using APIs from providers such as OpenAI, Anthropic, or Google, and open-source models.
• RAG: practical experience with retrieval-augmented generation pipelines, vector databases, hybrid search, and retrieval evaluation.
• Agentic systems: experience with frameworks such as LangChain, LlamaIndex, or LangGraph - tool use, multi-agent orchestration, and system-level prompt/context engineering.
• Model Context Protocol (MCP): practical experience building or integrating MCP-based tool servers.

Responsibilities

• Lead the design and implementation of gen ai and llm-powered features across CAST's deterministic data.
• Architect and develop scalable, fault-tolerant RAG systems over large enterprise codebases with vector databases and hybrid search.
• Build agentic systems for code understanding and modernization: tool use via MCP, multi-step reasoning, and multi-agent orchestration.
• Drive architectural consistency across AI services and reduce duplication by shared libraries and reusable components.

Skills

Gen AI features

RAG pipelines

Agentic systems

LangChain tooling

LLM integration

Tools

LangChain

LlamaIndex

LangGraph

Job description

Businesses move faster using CAST technology to understand, improve, and transform their software. Through semantic analysis of source code, CAST produces 3d maps and dashboards to navigate inside individual applications and across entire portfolios. This intelligence empowers executives and technology leaders to steer, speed, and report on initiatives such as technical debt, generai, modernization, and cloud. As the pioneer of the software intelligence field, CAST is trusted by the world's leading companies and governments, their consultancies and cloud providers. See it all castsoftware.com.

About the Job:

Working at CAST R&D means being an important part of a highly-talented, fast-paced, multicultural and agile team based in Paris (France) and Bangalore (India). The team builds a sophisticated source code analysis platform leveraging parsing, control flow, data flow and other mechanisms to fully understand the inner structure of the complex IT systems developed and used by Fortune 500companies.

As a senior ai engineer, you will lead the design and delivery of the ai capabilities that sit on top of this uniquely rich, deterministic software-intelligence data — building gen ai and agentic systems that help enterprises understand, modernize, and transform their code with precision and certainty. The ai squad currently spans several independent services (transaction/code summarization, graph rag-based knowledge retrieval, and a multi-agent orchestration layer); a central part of this role is bringing architectural coherence across them as the squad and its codebase mature.

• Lead the design and implementation of gen ai and llm-powered features that turn cast's deterministic code and architecture data into actionable insights for developers, architects, and it executives.
• Architect and develop scalable, fault-tolerant rag systems over large enterprise codebases vector databases, hybrid search, chunking, and retrieval evaluation integrated as shared, reusable components across microservices rather than duplicated per service.
• Design and build agentic systems for code understanding and modernization: tool use via the
• model context protocol (mcp) multi-step reasoning, and multi-agent orchestration, including handling long-horizon task failure modes with durable, resumable workflows.
• Lead efforts to optimize inference performance and cost in production, leveraging best practices in prompt/context engineering, caching, model selection, and latency/through put tuning.
• Establish evaluation, guardrails, and responsible-ai standards to measure and continuously improve accuracy, safety, and reliability — including building the squad's first retrieval/agent evaluation harness.
• Drive architectural consistency across the ai squad's services consolidate duplicated llm-provider, graph-database, and configuration logic into shared, well-tested internal libraries; introduce lightweight architecture-decision records to capture rationale as the platform evolves.
• Provide technical guidance and mentorship to engineers, fostering a culture of excellence, collaboration, and continuous learning within the team.

You will be working as an individual contributor, integrated in teams working on the ai and dashboard capabilities of the cast platform. Your teammates are located in india and in france. You will collaborate on writing and designing new features and improving existing ones. You will write unit tests and drive code reviews. You will participate in best-practices definition and technology watch. Depending on your will, skills and experience, you will have the opportunity to take technical lead on topics or projects.

While the domain of cast is a niche, the position will offer you the chance to work on software at the intersection of generative ai and deterministic software intelligence — a genuinely differentiated technical problem — focusing on technical and creative skills.

Profile Experience:

• 4+ years, with hands-on experience building and shipping ai/ml systems in production.
• Core ai/gen ai skills: building llm applications using apis from providers such as open ai, anthropic, or google, and open-source models.
• rag: practical experience with retrieval-augmented generation pipelines, vector databases(graph-native vector indexes a plus), hybrid search, and retrieval evaluation.
• Agentic systems: experience with frameworks such as lang chain, llama index, or lang graph - tool use, multi-agent orchestration, and system-level prompt/context engineering.
• Model context protocol (mcp): practical experience building or integrating mcp-based tool servers - this is core to how the squad exposes agent tooling, not a peripheral skill.