R Systems - Lead Engineer - LLM/RAG

R Systems; Bengaluru; On-site; INR 2,400,000 - 4,000,000; Full time

Job summary

R Systems is seeking an experienced software engineer to design and ship LLM-powered applications from Bangalore. You will work on end-to-end RAG systems, agentic workflows, and robust evaluation in production settings.

The role requires deep Python skills, experience with major LLM providers, and strong tooling fluency across LangChain, vector stores, and containerized deployments. Immediate joiner is preferred.

Qualifications

• 4+ years of software engineering experience, with 2+ years building and shipping LLM-powered applications in production.
• Strong Python skills - production-quality code, clean architecture, type-safe data modeling (Pydantic or equivalent), and testable design.
• Hands-on production experience with at least one major LLM provider - AWS Bedrock, Anthropic Claude, OpenAI GPT, Google Gemini, or equivalent - including tool/function calling and structured output.
• Demonstrated experience building RAG systems end to end - chunking, embeddings, vector databases (Qdrant, Pinecone, Weaviate, OpenSearch, or pgvector), retrieval, and re-ranking.
• Hands-on experience building agentic workflows - LLM agents that call tools, reason across multiple steps, and autonomously complete multi-stage tasks with appropriate safeguards.
• Experience with fine-tuning and adaptation workflows - LoRA, QLoRA, instruction tuning, or preference tuning - and with evaluating model outputs rigorously rather than by demo.
• Solid LLM tooling fluency - LangChain or LangGraph, LlamaIndex, DSPy, Hugging Face, and the judgment to pick the right tool rather than adopt reflexively.
• Prompt engineering craft - system prompts, few-shot examples, chain-of-thought, structured output schemas, and tool/function calling.
• Comfort working with unstructured data - extracting, classifying, and generating structured outputs from text-heavy inputs, including documents and forms.
• Cloud and deployment fluency - AWS preferred (including Bedrock), plus containerization (Docker), and familiarity with self-hosted LLM serving (vLLM, TGI, Ollama, or similar).
• Evaluation discipline - ability to write tests for non-deterministic systems, build golden sets, and reason about output quality at scale.
• Strong debugging instincts for LLM-specific failure modes - hallucinations, retrieval gaps, prompt drift, and latency spikes.

Skills

Python programming

LLM fundamentals

RAG systems

LLM agents

Prompt engineering

LLM evaluation

Unstructured data handling

Debugging LLMS

Testable design

Production-quality code

Tools

AWS Bedrock

LangChain/LangGraph

LlamaIndex

DSPy

Hugging Face

Docker

Qdrant

Pinecone

Weaviate

OpenSearch

vLLM

Ollama

Job description

Urgent opening with R systems.

Required Qualifications

Job location : Bangalore (Mandatory).

Work from Office : Yes.

Looking for Immediate joiner only.

Exp level : 6 - 9 Yrs.

Core Experience

• 4+ years of software engineering experience, with 2+ years building and shipping LLM-powered applications in production.
• Strong Python skills - production-quality code, clean architecture, type-safe data modeling (Pydantic or equivalent), and testable design.
• Hands-on production experience with at least one major LLM provider - AWS Bedrock, Anthropic Claude, OpenAI GPT, Google Gemini, or equivalent - including tool/function calling and structured output.
• Demonstrated experience building RAG systems end to end - chunking, embeddings, vector databases (Qdrant, Pinecone, Weaviate, OpenSearch, or pgvector), retrieval, and re-ranking.
• Hands-on experience building agentic workflows - LLM agents that call tools, reason across multiple steps, and autonomously complete multi-stage tasks with appropriate safeguards.
• Experience with fine-tuning and adaptation workflows - LoRA, QLoRA, instruction tuning, or preference tuning - and with evaluating model outputs rigorously rather than by demo.
• Solid LLM tooling fluency - LangChain or LangGraph, LlamaIndex, DSPy, Hugging Face, and the judgment to pick the right tool rather than adopt reflexively.
• Prompt engineering craft - system prompts, few-shot examples, chain-of-thought, structured output schemas, and tool/function calling.
• Comfort working with unstructured data - extracting, classifying, and generating structured outputs from text-heavy inputs, including documents and forms.
• Cloud and deployment fluency - AWS preferred (including Bedrock), plus containerization (Docker), and familiarity with self-hosted LLM serving (vLLM, TGI, Ollama, or similar).
• Evaluation discipline - ability to write tests for non-deterministic systems, build golden sets, and reason about output quality at scale.
• Strong debugging instincts for LLM-specific failure modes - hallucinations, retrieval gaps, prompt drift, and latency spikes.

Engineering & Platform Skills

• 4+ years of software engineering experience, with 2+ years building and shipping LLM-powered applications in production.
• Strong Python skills - production-quality code, clean architecture, type-safe data modeling (Pydantic or equivalent), and testable design.
• Hands-on production experience with at least one major LLM provider - AWS Bedrock, Anthropic Claude, OpenAI GPT, Google Gemini, or equivalent - including tool/function calling and structured output.
• Demonstrated experience building RAG systems end to end - chunking, embeddings, vector databases (Qdrant, Pinecone, Weaviate, OpenSearch, or pgvector), retrieval, and re-ranking.
• Hands-on experience building agentic workflows - LLM agents that call tools, reason across multiple steps, and autonomously complete multi-stage tasks with appropriate safeguards.
• Experience with fine-tuning and adaptation workflows - LoRA, QLoRA, instruction tuning, or preference tuning - and with evaluating model outputs rigorously rather than by demo.
• Solid LLM tooling fluency - LangChain or LangGraph, LlamaIndex, DSPy, Hugging Face, and the judgment to pick the right tool rather than adopt reflexively.
• Prompt engineering craft - system prompts, few-shot examples, chain-of-thought, structured output schemas, and tool/function calling.
• Comfort working with unstructured data - extracting, classifying, and generating structured outputs from text-heavy inputs, including documents and forms.
• Cloud and deployment fluency - AWS preferred (including Bedrock), plus containerization (Docker), and familiarity with self-hosted LLM serving (vLLM, TGI, Ollama, or similar).
• Evaluation discipline - ability to write tests for non-deterministic systems, build golden sets, and reason about output quality at scale.
• Strong debugging instincts for LLM-specific failure modes - hallucinations, retrieval gaps, prompt drift, and latency spikes.

Security & Regulated-Industry Awareness

• Working awareness of LLM security concerns - prompt injection, data exfiltration, output filtering, and PII handling.
• Respect for data sensitivity and an instinct to keep borrower PII and financial data inside Lendistry's trust boundary by default.

Preferred Qualifications

• Experience in fintech, lending, banking, healthcare, or another regulated or data-sensitive industry.
• Familiarity with document understanding models (LayoutLM, Donut, Nougat) and modern OCR tooling (Textract, Tesseract, or equivalents).
• Background in NLP tasks such as named entity recognition, classification, or semantic similarity.
• Experience with LLaMA-family open-weight models and self-hosted serving via Ollama or vLLM.
• Experience contributing to a shared AI platform (prompt registry, evaluation harness, routing layer) used by multiple product teams.

(ref:hirist.tech)