LLM Engineer - RAG/Prompt Engineering

Description:Job Title: LLM Engineer (Mid / Senior)Locations: Gurugram (Sector 54) / Chennai (T. Nagar) Hybrid (4 Days In-Office)Experience: 3+ YearsRole Summary:We are an agentic AI platform for financial services, backed by Y Combinator and world-class investors. As a Mid-to-Senior LLM Engineer, you will be pivotal in evolving our V1 platform into a scalable, enterprise-grade system for global banks and financial institutions. We are looking for engineers who have a "founder mindset," a strong bias for action, and deep technical expertise in orchestrating complex AI workflows.Key Responsibilities:- Model Orchestration: Design and manage complex workflows across both open-source and closed-source LLMs.- RAG & Knowledge Graphs: Build robust pipelines for data cleaning and transformation for RAG; research and implement SOTA approaches like Graph RAG.- Evaluation & Optimization: Develop and run multi-rubric evals for text classification, sentiment analysis, and STT; implement inference-time reasoning to boost model accuracy.- Data Mastery: Clean, prune, and collect high-quality datasets to minimize hallucination rates and optimize recall and performance.- Scalable Infrastructure: Take ownership of moving AI features from prototype to a scalable, robust production environment.Required Qualifications:- Technical Pedigree: 3+ years of experience with Machine Learning, NLP, and Prompt Engineering.- Proven Impact: Experience building systems from 0 to 1 or 1 to 10 (typically in early or growth-stage startups.- Hands-on AI Experience: Mandatory experience in optimizing prompts, building RAG pipelines, and running rigorous model evals.- Action-Oriented: A track record of thoughtful work evidenced by side projects, open-source contributions, or technical blogs rather than just a CV.- Cultural Alignment: Deep resonance with our values: "Doers over Talkers," 360-degree feedback, and human-centric design.Our Core Values:- Product Ownership: Everyone, including engineers, engages in customer outreach and support to identify and build the right features.- Flat Hierarchy: No one is too senior for any taskour CTO codes daily, and everyone handles customer support on a rotation.- Empathy First: We build for humans. Kindness and empathy are core requirements for every team member. (ref:hirist.tech)