Junior RAG And LLM Integration Developer

What role will you play in the team

You will serve as a foundational technical team member specializing in deploying Large Language Models and Retrieval-Augmented Generation pipelines. Your primary focus will be bridging unstructured enterprise data repositories with modern transformer architectures to build context-aware cognitive search capabilities that drastically reduce hallucination rates and improve retrieval accuracy for downstream AI applications.

What you will do

You will write production-grade Python code to parse complex documents, chunk vector embeddings, and interface directly with proprietary APIs. Working closely with senior machine learning engineers, you will benchmark latency, optimize vector database queries, and establish automated evaluation frameworks to continuously monitor the factual correctness and semantic relevance of generated outputs across multiple enterprise verticals.

Key responsibility

• Design, implement, and maintain advanced Retrieval-Augmented Generation (RAG) pipelines using Python, LangChain, and LlamaIndex for enterprise search applications.
• Integrate OpenAI API, Claude API, and open-source Large Language Models (LLMs) into existing microservices architecture via secure RESTful endpoints and webhooks.
• Manage high-performance vector databases such as Pinecone, Milvus, and PostgreSQL with pgvector for efficient similarity search operations.
• Fine-tune open-source embedding models and implement re-ranking mechanisms to boost precision and recall metrics in domain-specific corpora.
• 
  Collaborate with data engineering teams to establish automated ETL pipelines that ingest, clean, and vectorize unstructured PDF, HTML, and database text.
• Implement rigorous prompt engineering strategies, including few-shot learning, chain-of-thought prompting, and guardrails to filter toxic or irrelevant completions.
• Monitor production LLM application performance, tracking token consumption, latency thresholds, and API rate limits to optimize operational expenditure.

Required Qualification and Skills

• Bachelor's degree in Computer Science, Artificial Intelligence, Data Science, or a related quantitative discipline with 0 to 1 years of hands-on software development experience.
• Strong proficiency in programming languages, specifically Python, with deep familiarity with asynchronous programming and object-oriented design principles.
• Demonstrated project experience or academic background working with Large Language Models, OpenAI API, Claude API, and prompt engineering frameworks.
• Familiarity with vector database management systems (Pinecone, Qdrant, Chroma, or PostgreSQL) and embedding generation techniques.
• Understanding of machine learning frameworks and libraries including scikit-learn, PyTorch, and Hugging Face Transformers.
• Experience utilizing version control systems like Git, containerization tools like Docker, and deploying applications via cloud infrastructure.
• Excellent analytical mindset, rapid prototyping capability, and strong problem-solving skills for debugging complex natural language processing pipelines.

Benefits Included

• Comprehensive health insurance coverage extending to immediate family members with cashless hospitalization options across major hospital networks.
• Flexible working hours and hybrid work arrangement policies designed to promote a healthy work-life integration.
• Dedicated annual learning and development stipend for pursuing professional certifications, online courses, and attending industry AI conferences.
• Generous paid time off, casual leave allocations, and mental health wellness days provided throughout the calendar year.
• Performance-linked annual bonuses, quarterly reward recognitions, and structured career advancement pathways within the artificial intelligence division.

A Day in the Life

Your day begins by reviewing real-time monitoring dashboards to assess the latency and token usage of deployed RAG pipelines. You then spend the morning writing custom Python scripts to parse new technical documentation formats and ingest them into a vector database. After collaborating during a daily sprint standup, you dedicate your afternoon to experimenting with advanced prompt engineering techniques and fine-tuning retrieval parameters to resolve edge-case user queries before wrapping up your tasks by pushing clean code updates to the staging repository.