Ai Engineer

## About the Company Supaboard.ai is an AI-powered business intelligence platform that unifies data from warehouses, lakes, and over 110 third-party sources to deliver instant dashboards, KPIs, and visual reports through plain English queries—no SQL or spreadsheets required. Industry-trained AI analysts provide precise, domain-specific insights for sectors like e-commerce, fintech, and SaaS, enabling secure, real-time decision-making for teams of any size. We believe in lean teams delivering high impact, and Supaboard's small, focused crew is building a next-generation analytics engine that fuses automated data modeling with specialized AI capabilities. This approach empowers companies to bypass the overhead of large data science teams while achieving fast, scalable analytics. Supaboard's mission is to transform scattered data silos into a trusted, unified hub for actionable insights that drive business growth. Engineers here shape the future of AI-driven analytics, with direct influence on product features used by real customers daily. ## About the Role Position: AI Engineer Location: Bangalore (On-site, 5 days a week) ## Key Responsibilities - Fine-tune open-source LLMs/SLMs for use cases like summarization, classification, report generation, and retrieval workflows using Hugging Face or PyTorch. - Build and maintain RAG pipelines with vector databases (e.g., Pinecone, Weaviate, Chroma) for efficient data retrieval and embedding-based search. - Design prompt workflows, chains, and agent-style orchestration for reliable, user-facing AI features. - Integrate ML/NLP models (e.g., classification, extraction, summarization) into analytics tools. - Collaborate on backend services (Python + REST/Graph APIs) to deploy AI in production SaaS environments. - Monitor model performance with metrics, logging (e.g., W&B), error analysis, and iterative improvements. - Work with product/data teams to turn user queries into robust AI/ML pipelines for real-world business data. - Ensure code quality, testing, documentation, scalability, and secure data handling. ## Must-Have Skills - Experience fine-tuning/training open-source LLMs/SLMs (Hugging Face, PyTorch). - Hands-on with vector databases and retrieval pipelines. - Strong grasp of RAG, embeddings, prompt engineering, tuning, chaining/agents. - Proficiency building ML/NLP models (classification, summarization, extraction). - Expert Python skills; backend frameworks/SOA a plus. - Ability to build production-grade AI features for SaaS/analytics. - Cloud familiarity (AWS/GCP) preferred. - Analytical mindset for debugging and owning end-to-end features. ## Nice-to-Have Skills - Multi-model pipelines, agents, or tool-calling (e.g., retrieval → LLM → post-processing). - PEFT techniques (LoRA, quantization) for LLM optimization. - Backend frameworks (FastAPI, Node.js) and AI integration in web/dashboards. - Ingesting structured/semi-structured data for BI features. ## Why Join Supaboard.ai - Core role in India's leading AI analytics startup, building production features for real users. - Shape model fine-tuning, retrieval architecture, and product roadmap with high ownership. - Rapid growth in a lean, collaborative environment focused on curiosity and innovation.