Enterprise AI Specialist
Explicitly requires vibe coding skills and agentic workflows; uses AI agents and LLM orchestration for rapid prototyping and delivery.
About the Role
Enterprise AI Specialist at Finastra responsible for driving AI adoption across applications by translating business intent into LLM-orchestrated solutions and agentic workflows. The role focuses on prompt engineering, context/RAG optimization, prompt evaluation, safety hardening, and enabling stakeholders through workshops and rapid prototyping.
Job Description
Role
Finastra is hiring an Enterprise AI Specialist (Prompt Engineer & Agentic Systems) reporting to the Director of Transformation in Bengaluru (hybrid). The specialist will act as the primary catalyst for AI adoption across the application portfolio, translating non-technical business intent into production-ready AI-driven behaviors using LLM orchestration and agentic workflows.
Key Responsibilities
- Design, test, version-control and maintain complex system prompts to ensure consistent, high-quality outputs.
- Lead training and enablement workshops for application owners to improve prompt literacy and business outcomes.
- Develop prompt architectures and multi-turn system instructions using patterns such as Chain-of-Thought, Tree-of-Thought and ReAct.
- Establish PromptOps practices, including a “Golden Dataset” for regression testing and prompt evaluation.
- Optimize RAG pipelines: chunking, indexing, and context injection for effective retrieval-augmented generation.
- Audit and optimize prompts for token and cost efficiency while preserving accuracy.
- Implement safety and adversarial hardening (guardrail prompts) to prevent prompt injection and data exfiltration.
- Act as a technical translator (vibe-to-code) to rapidly prototype and iterate application features in natural language.
- Build, debug and deploy agentic workflows using AI coding environments to accelerate full-stack delivery.
- Read, audit, and debug AI-generated code to ensure security and scalability.
Requirements
- 2–6 years relevant experience.
- Deep familiarity with frontier LLMs and their differences (explicitly mentions OpenAI, Claude, Anthropic, Google models and open-source models such as Llama).
- Mastery of advanced prompting techniques (Few-Shot, Zero-Shot, Dynamic Prompting) and output structuring (JSON, YAML, Pydantic).
- Experience with programmatic prompting/orchestration frameworks (e.g., LangChain, DSPy, Semantic Kernel).
- Experience optimizing and operating RAG pipelines and connecting models to internal data sources (Jira, GitHub, SQL databases).
- Practical experience with agentic systems and AI coding environments (e.g., Cursor, Windsurf, Replit Agent).
- Strong verification and debugging skills to validate AI outputs for security and scalability.
- Ability to facilitate workshops and communicate technical concepts to non-technical stakeholders.
Preferred
- Strong intuition for AI-first system architecture and rapid iteration with AI agents.
- Pragmatic judgment on when to hand-code components versus using vibe-coded prototypes.
Benefits (summary)
Finastra lists hybrid working arrangements, unlimited vacation (location-dependent), paid time off for voting/bereavement/sick leave, Employee Assistance Program (confidential therapy), wellbeing programs, medical/life/disability insurance, retirement plan, paid volunteer time and donation matching, DE&I groups, learning and career development tools, and a global recognition program.