Company Description Dimensionless Technologies is a global IT consultancy and AI services company that delivers AI-based solutions across a wide range of industries. Founded in 2016 in Mumbai, India, the company is driven by innovation and a focus on transforming businesses through artificial intelligence. Its team of data scientists, machine learning engineers, and software developers builds bespoke solutions such as security automation, fraud detection, and natural language processing.

With offices in Mumbai and New Jersey and a strong presence in both the US and India, Dimensionless serves a diverse international clientele. The organization emphasizes operational efficiency, growth, and cutting-edge AI applications for its clients.

About the Role:

We are looking for an AI Engineer with 3–4 years of hands-on experience building production-grade Retrieval-Augmented Generation (RAG) systems and LLM-powered applications. You will design, build, and scale AI solutions, working closely with product, data, and platform teams to take ideas from prototype to production. Familiarity with emerging protocols like Model Context Protocol (MCP) and Agent-to-Agent (A2A) communication is a strong plus, as we are actively building agentic, tool-using AI systems.

What you will do

- Design, build, and maintain RAG pipelines, including chunking strategies, embedding generation, vector search, and retrieval optimization
- Develop and fine-tune LLM-based applications for tasks such as summarization, Q&A;, extraction, classification, and conversational agents
- Build and integrate tool-using agents leveraging protocols like MCP for standardized tool/context access and A2A for multi-agent communication and orchestration
- Evaluate and select appropriate LLMs (open-source and proprietary), embedding models, and vector databases based on use case requirements
- Implement prompt engineering, prompt chaining, and structured output strategies to improve reliability and accuracy




- Build evaluation frameworks to measure RAG/LLM output quality (relevance, faithfulness, hallucination rate, latency, cost)
- Optimize retrieval and generation pipelines for latency, cost, and accuracy at scale
- Collaborate with backend engineers to integrate AI features into production applications via APIs and microservices
- Implement guardrails, safety checks, and monitoring for LLM outputs in production
- Stay current with the fast-evolving LLM/agentic AI ecosystem and bring in relevant advances

Required Qualifications

Must-Have:

- 3–4 years of professional software engineering experience, with at least 1.5–2 years focused on LLM/RAG-based systems
- Strong Python skills; experience with frameworks like LangChain, LlamaIndex, Haystack, or similar
- Practical experience with vector databases (Pinecone, Weaviate, Qdrant, Milvus, pgvector, etc.)
- Experience working with LLM APIs (OpenAI, Anthropic Claude, open-source models via Hugging Face, etc.)
- Solid understanding of embedding models, semantic search, and retrieval techniques (hybrid search, re-ranking, chunking strategies)
- Experience with prompt engineering and structured output generation (function calling, JSON schemas, etc.)
- Understanding of evaluation methodologies for LLM outputs (accuracy, hallucination detection, relevance scoring)
- Familiarity with REST APIs, containerization (Docker), and basic cloud deployment (AWS/GCP/Azure)

Valuable to Have:

- Working knowledge of Model Context Protocol (MCP) for connecting LLMs with external tools, data sources, and context
- Understanding of Agent-to-Agent (A2A) protocols and multi-agent orchestration frameworks (AutoGen, CrewAI, LangGraph, etc.)
- Experience fine-tuning or instruction-tuning open-source LLMs (LoRA, QLoRA, PEFT)
- Exposure to LLMOps/MLOps practices — model monitoring, versioning, A/B testing of prompts/models
- Experience with streaming architectures and real-time inference pipelines
- Familiarity with responsible AI practices — bias mitigation, safety guardrails, PII handling

📌 AI Engineer (India)
🏢 Dimensionless Technologies
📍 India