AIML Engineer

Job Title: AI Engineer - LLMs, RAG, Agentic AI & Automation

Location : Noida (Hybrid)

Experience : 0 2 years

Employment Type: Full-time

About the Role

• Were looking for a curious, hands-on AI Engineer who wants to work at the intersection of Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and Agentic AI.
• You will design, prototype, and productionize multi-agent systems that integrate reasoning, retrieval, automation, and real-world workflows.
• Alongside this, youll also apply your foundation in traditional machine learning and deep learning to build, fine-tune, and evaluate models bringing together classical AI principles with cutting-edge generative techniques.
• This role is ideal for candidates who have a research-oriented mindset and are passionate about building systems that move from prototype to real-world deployment.

Key Responsibilities : -

LLM & RAG Engineering : -

• Build and optimize retrieval-augmented generation (RAG) pipelines (chunking, embeddings, re-ranking, retrieval, generation).
• Work with frameworks like LangChain , LlamaIndex , and vector databases (Milvus, Qdrant , Weaviate , Pinecone).
• Experiment with context-aware retrieval and hybrid search methods (semantic + keyword). Agentic AI Development
• Design and implement multi-agent architectures using frameworks like CrewAI , LangGraph , or custom FastAPI orchestration.
• Enable tool-use, planning, and reasoning capabilities for autonomous systems.
• Integrate human-in-the-loop (HITL) feedback for iterative improvement.
• Fine-Tuning & Adaptation
• Fine-tune open-weight LLMs using LoRA , PEFT, DPO, and instruction-tuning pipelines.
• Collect and curate high-quality datasets for supervised and reinforcement-style fine-tuning.
• Evaluate performance across tasks using custom and standard benchmarks.

Browser & Workflow Automation : -

• Develop browser automation and data-gathering agents using Playwright, Selenium, or Puppeteer.
• Integrate browser-based tools into agent workflows for real-time data retrieval and task execution.
• Traditional ML/DL Applications
• Implement, train, and evaluate models for classification, regression, and clustering tasks using frameworks like scikit-learn, PyTorch , or TensorFlow.
• Work with feature engineering, model selection, and performance tuning pipelines.
• Deploy models as REST APIs or microservices.

Research & Experimentation : -

• Conduct ablation studies, write experiment reports, and suggest improvements.
• Contribute to internal technical notes, blogs, or open-source work.

Technical Skills : -

Required : -

• Python (strong foundation)
• Experience with LLM APIs ( OpenAI , Anthropic, Gemini, etc.)
• Understanding of Transformer architecture, attention, embeddings
• Familiarity with LangChain , LlamaIndex , or custom RAG setups
• Experience with vector databases (Milvus, Qdrant , Weaviate , Chroma, Pinecone)
• Basic knowledge of traditional ML algorithms (logistic regression, random forest, SVM, clustering)
• Working knowledge of PyTorch or TensorFlow
• Multi-agent frameworks ( CrewAI , ADK, LangGraph , AutoGen )
• Browser automation (Playwright, Puppeteer)
• FastAPI / Flask for API development

Desirable (Good to Have):

• Cloud deployment (GCP, AWS, Azure, Modal, Vercel )
• Familiarity with Git, Docker, and CI/CD pipelines
• Prior research publications, technical blog posts, or open-source contributions

Soft Skills : -

• Analytical thinking and attention to detail
• Research curiosity and eagerness to learn new architectures quickly
• Ability to work in agile, fast-iteration environments
• Good written and verbal communication skills
• Ownership mentality with a bias toward experimentation

Preferred Background : -

• B.Tech / B.E. / M.Tech / M.S. in Computer Science, IT, AI, or related field
• Internship or project experience with ML/DL or NLP
• Participation in hackathons, Kaggle, or research work is a strong plus
• Demonstrated interest in LLMs, automation, or cognitive AI systems

Skills : -

• Python, LLM APIs, OpenAI, Anthropic, Gemini, Transformer architecture, attention, embeddings, LangChain, LlamaIndex, custom RAG setups, vector databases, Milvus, Qdrant, Weaviate, Chroma, Pinecone, traditional ML algorithms, logistic regression, random forest, SVM, clustering, PyTorch, TensorFlow, multi-agent frameworks, CrewAI, ADK, LangGraph, AutoGen, browser automation, Playwright, Puppeteer, FastAPI, Flask