LLM Engineer (Large Language Models)

TymblHub; Bengaluru; On-site; INR 1,800,000 - 2,800,000; Full time

Job summary

Fospe Software Pvt. Ltd. in Bengaluru seeks an experienced LLM Engineer to architect, fine-tune, and deploy open-source LLMs and RAG pipelines for enterprise platforms.

You will work with our AI research team to embed deterministic cognitive capabilities and high-throughput semantic reasoning into our vertical SaaS products. You will collaborate with backend engineers to expose scalable interfaces, optimize for low-latency inference, and apply guardrails, safety filters, and benchmarking.

Qualifications

• 3+ years of professional ML/DL experience with 1+ years in LLM engineering.
• Experience with open-source foundation models and quantization techniques is required.

Responsibilities

• Architect, fine-tune, and quantize LLMs (Llama 3, Mistral, Qwen, DeepSeek) for enterprise workloads.
• Design RAG architectures with hybrid dense/sparse search and reranking.

Skills

Python

PyTorch

Hugging Face

vLLM

LangChain

LlamaIndex

CUDA

Async programming

Education

Bachelor's or Master's in CS/AI/Data Science

Tools

Docker

Kubernetes

Linux

CUDA Toolkit

Job description

# LLM Engineer (Large Language Models)Fospe Software Pvt. LtdFPosted on September 3, 2026---Experience3 - 8 yrsJob LocationBengaluru, IndiaVacancy1DesignationLLM EngineerJob TypeNot specified---## Job Description### Job SummaryArchitect, fine-tune, and deploy large language models (LLMs) and advanced retrieval-augmented generation (RAG) search pipelines for enterprise platforms.As an LLM Engineer at Fospe, you will be at the forefront of designing, fine-tuning, and operationalizing cutting-edge Large Language Models and Retrieval-Augmented Generation (RAG) pipelines. You will collaborate directly with our AI research team to embed deterministic cognitive capabilities and high-throughput semantic reasoning into our vertical enterprise SaaS products.### Responsibilities* Architect, fine-tune, and quantize open-source foundation models (Llama 3, Mistral, Qwen, DeepSeek) for domain-specific enterprise workloads.* Design high-accuracy Retrieval-Augmented Generation (RAG) architectures with hybrid dense/sparse search, reranking, and semantic chunking.* Build low-latency streaming inference pipelines utilizing vLLM, TensorRT-LLM, and Triton Inference Server.* Implement guardrail validation, safety filters, prompt routing, and automated hallucination benchmarking.* Collaborate with backend engineers to expose scalable gRPC and REST interfaces for seamless frontend consumption.### Requirements* Bachelor s or Master s degree in Computer Science, Artificial Intelligence, Data Science, or related quantitative field.* 3+ years of professional experience in machine learning and deep learning, with 1+ years dedicated to LLM engineering.* Deep hands-on expertise with PyTorch, Hugging Face Transformers, vLLM, and LangChain/LlamaIndex.* Proven experience with vector databases such as Pinecone, Qdrant, Milvus, or pgvector.* Strong proficiency in Python, asynchronous programming, Docker containerization, and Linux/CUDA environments.* Demonstrated understanding of model optimization techniques including LoRA, QLoRA, AWQ, and GPTQ quantization.### Preferred Qualifications* Experience with multi-agent orchestration frameworks (LangGraph, CrewAI, AutoGen).* Contributions to open-source AI libraries or published research in NLP/LLMs.* Experience deploying models on Kubernetes and cloud infrastructure (AWS/GCP/Azure GPU instances).### Tech Stack* Python* PyTorch* Hugging Face* vLLM* LangChain* LlamaIndex* Qdrant* Docker* Kubernetes* CUDA* FastAPIDisclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.