GenAI Engineer

## About the Role We're seeking a GenAI Agent Engineer to join the ServiceHive team as a core agent builder. This is an individual contributor (IC) role where you'll take end-to-end ownership of designing, building, and deploying intelligent AI agents that automate customer interactions across platforms like websites, WhatsApp, Instagram, and other digital touchpoints. You'll play a key role in building AI that helps businesses automate conversations, engage users, and convert leads 24/7. Your Mission: You'll be the primary hands-on builder of our AI agents - designing conversation flows, training models, implementing agent logic, and continuously improving agent quality. This is a high-ownership role where you'll ship agent features independently. ## About ServiceHive At ServiceHive, we are building the future of AI-powered customer engagement. Our platform enables businesses to automate conversations across multiple channels—helping them capture leads, support customers, and scale operations effortlessly. We believe in creating practical, high-impact AI solutions that businesses can rely on daily. As part of our team, you'll work on real-world problems, ship meaningful features, and contribute directly to building a product that transforms how businesses communicate. ## Experience Level - 2-3 years of professional experience with strong focus on GenAI, LLMs, and AI agent development - 3-5 years of experience in backend development or machine learning ## Key Responsibilities ### Core Agent Development (Primary Focus - 80% of time) - Own the day-to-day development of conversational AI agents - this is your primary responsibility - Design and implement agent architectures, conversation flows, and decision logic - Build and fine-tune LLM-based agents using frameworks like LangChain, LlamaIndex, or custom implementations - Implement prompt engineering strategies to optimize agent responses for different use cases - Train and fine-tune models for intent recognition, entity extraction, and sentiment analysis - Build conversation state management, memory systems, and context handling from scratch - Implement multi-turn conversation logic and dialogue management - Design and implement agent tools, function calling, and external API integrations - Optimize agent performance, response quality, and conversation success rates - Conduct experiments and A/B tests to improve agent effectiveness - Monitor agent performance and iterate based on conversation analytics ### Backend Integration (15% of time) - Integrate agents with backend APIs and services (with support from Senior Backend Engineer) - Implement features like lead capture, FAQ responses, and conversation routing - Write unit tests and integration tests for agent logic - Debug and fix issues in agent systems ### Collaboration & Growth (5% of time) - Collaborate with the Data Scientist on complex agent challenges and new research directions - Participate in code reviews and share knowledge with the team - Stay updated with latest LLMs, agent frameworks, and GenAI research - Contribute ideas for improving agent capabilities and conversation quality - Document agent architectures, prompts, and technical decisions ## Required Qualifications ### Must-Have Skills (Non-Negotiable) #### GenAI & Agent Development (Core Requirements) - LLM Experience: Strong hands-on experience working with LLMs (OpenAI GPT-4, Claude, Gemini, or open-source models) - Agent Frameworks: Solid experience with at least one agent framework (LangChain, LlamaIndex, AutoGen, CrewAI, Haystack, or similar) - Prompt Engineering: Proven ability to design, test, and optimize prompts for different use cases - Conversational AI: Experience building chatbots, conversational agents, or dialogue systems - Agent Patterns: Understanding of agent architectures (ReAct, Chain-of-Thought, Tool Use, Memory, etc.) - NLP Fundamentals: Understanding of NLP concepts (intent recognition, entity extraction, sentiment analysis, text classification) #### Programming & Development - Python: Strong proficiency in Python (primary language for agent development) - LLM APIs: Experience integrating LLM APIs (OpenAI, Anthropic, Cohere, HuggingFace, etc.) - Backend Basics: Basic understanding of APIs, databases, and backend services - Version Control: Proficient with Git and collaborative development workflows - Testing: Experience testing agent outputs, conversation flows, and LLM responses #### Proven Track Record - Portfolio Required: Demonstrable projects or work experience building AI agents, chatbots, or LLM applications - Examples could include: - Production chatbots or conversational AI systems - LLM-based applications or tools - Personal projects using LangChain, LlamaIndex, or similar frameworks - Contributions to open-source agent projects - Research or experimentation with agent architectures - Ability to discuss technical decisions, challenges, and learnings from past agent projects #### Soft Skills - High Ownership: Comfortable taking full ownership of agent development and shipping features independently - Self-Motivated: Ability to work independently and drive agent improvements without constant direction - Problem-Solving: Strong analytical thinking and debugging skills for complex agent behaviors - Communication: Clear communication about agent performance, technical decisions, and blockers - Experimentation Mindset: Comfortable running experiments, analyzing results, and iterating quickly - Attention to Detail: Obsessive about conversation quality and user experience - Curiosity: Passionate about GenAI, LLMs, and staying on top of the latest research ## Strong Plus (Highly Desired) - Fine-tuning Experience: Experience fine-tuning LLMs (GPT-3.5, Llama, Mistral, etc.) for specific use cases - Multi-Agent Systems: Experience building multi-agent systems or agent orchestration - RAG Expertise: Strong understanding and implementation of RAG (Retrieval Augmented Generation) patterns - Vector Databases: Hands-on experience with vector stores (Pinecone, Weaviate, Chroma, Qdrant) for agent memory - Evaluation Frameworks: Experience