kantiv - Data Scientist

Remote - India1w ago

Remote Junior APAC Artificial Intelligence Software Data Scientist Python Vector Documentation REST Pandas SQL pgvector Elasticsearch Pinecone Weaviate Docker B2B

Requirements

• Experience: 1–3 years of experience in AI engineering, machine learning engineering, software engineering, or a related field. Strong internships and hands-on projects involving AI applications are also valuable. • Programming: Proficiency in Python and the ability to write clean, modular, maintainable code. • LLM Fundamentals: Understanding of how LLMs work at an application level, including prompting, context windows, token limits, structured outputs, and common failure modes. • RAG & Embeddings: Familiarity with embeddings, vector search, document retrieval, and the fundamentals of building RAG-based applications. • APIs & Backend Fundamentals: Experience working with REST APIs, JSON, asynchronous operations, and integrating external services into applications. • Data Handling: Ability to work with structured and unstructured data, using tools such as Pandas and SQL where appropriate. • Testing & Debugging: Ability to test application behaviour systematically, investigate failures, and improve the reliability of AI-driven workflows. • Problem-Solving: Strong analytical thinking and the ability to break down ambiguous problems into practical engineering solutions. • Collaboration: Clear communication skills, openness to feedback, and the ability to work effectively with engineers and product stakeholders. • EXPERIENCE WE’D BE PARTICULARLY EXCITED ABOUT • Production LLM Applications: Experience building and deploying LLM-powered applications beyond simple prompt-based prototypes. • Agentic Systems: Hands-on experimentation with AI agents, tool calling, function calling, or multi-step reasoning workflows. • AI Frameworks: Familiarity with frameworks and libraries such as LangChain, LangGraph, LlamaIndex, or equivalent tools. • Vector Databases & Search: Experience with vector databases or search technologies such as pgvector, Pinecone, Weaviate, or Elasticsearch. • Evaluation & Observability: Exposure to LLM evaluation frameworks, tracing, monitoring, and tools for debugging AI workflows. • Cloud & Deployment: Familiarity with Docker, cloud platforms, deployment workflows, or serving AI applications in production. • Document Intelligence: Experience extracting information from PDFs, long-form documents, or complex business records. • B2B SaaS & Enterprise AI: Experience building AI capabilities for business users, particularly where accuracy, data security, and reliability matter. • AEC & Revenue Operations: Interest in applying AI to proposals, project information, client relationships, marketing, and revenue operations in the AEC industry.

Responsibilities

• AI Application Development: Build and improve AI-powered applications using LLMs, foundation models, and modern AI APIs to solve real-world product problems. • LLM Integration: Integrate commercial and open-source language models into applications, experimenting with prompting strategies, structured outputs, tool calling, and model selection. • RAG & Knowledge Systems: Develop retrieval-augmented generation (RAG) pipelines to help AI systems retrieve and reason over relevant information from documents, databases, and other unstructured data sources. • AI Agents & Workflows: Build and experiment with AI agents and multi-step workflows that use tools, access relevant context, and execute tasks with appropriate controls. • Data Processing & Retrieval: Develop pipelines for document ingestion, text extraction, chunking, embeddings, metadata management, and semantic search. • Evaluation & Reliability: Evaluate AI outputs for accuracy, relevance, consistency, and groundedness. Build test datasets and evaluation workflows to identify failure cases and improve system performance. • Backend Integration: Collaborate with backend engineers to integrate AI capabilities into APIs and existing product workflows, ensuring maintainability and reliability. • Performance & Cost Optimisation: Experiment with model selection, caching, retrieval strategies, and prompt optimisation to improve latency, reliability, and inference costs. • Monitoring & Debugging: Investigate issues in AI pipelines, troubleshoot unexpected model behaviour, and help monitor the quality and performance of AI-powered features. • Experimentation & Collaboration: Stay current with emerging AI techniques, evaluate relevant tools and frameworks, and work with product and engineering teams to translate customer problems into practical AI solutions.