Key Responsibilities

Design, develop, and deploy AI-driven applications and intelligent software solutions.
Build and optimize machine learning and deep learning models for production environments.
Develop applications leveraging Large Language Models (LLMs), Generative AI, and Natural Language Processing (NLP).
Fine-tune foundation models using domain-specific datasets and implement Retrieval-Augmented Generation (RAG) pipelines.
Develop AI services and RESTful APIs for seamless integration with enterprise applications.
Build scalable backend systems to support AI inference and model serving.
Design data pipelines for model training, evaluation, and continuous improvement.
Implement prompt engineering techniques to improve LLM performance and response quality.
Optimize model accuracy, latency, scalability, and infrastructure utilization.
Collaborate with software engineers, product managers, and data scientists throughout the product lifecycle.
Deploy AI models using containerization and orchestration technologies.
Implement MLOps practices for automated model training,



deployment, monitoring, and version control.
Evaluate emerging AI technologies and recommend suitable solutions for business use cases.
Ensure AI applications meet security, governance, and responsible AI standards.
Troubleshoot production issues and continuously improve system performance.

Required Skills

Strong programming skills in Python.
Experience with machine learning frameworks such as PyTorch, TensorFlow, or Scikit-learn.
Hands-on experience with Large Language Models (LLMs) and Generative AI technologies.
Solid understanding of NLP, Transformers, Embeddings, and Vector Databases.
Experience building RAG-based applications.
Knowledge of prompt engineering and LLM optimization techniques.
Experience with LangChain, LangGraph, LlamaIndex, or similar AI orchestration frameworks.
Experience developing REST APIs using FastAPI or Flask.
Strong understanding of data structures, algorithms, and

📌 Prismforce Pvt Ltd (Pune)
🏢 Prismforce
📍 Pune