Join us as a Custom Software Engineer, focused on advanced application engineering. This full-time role offers a unique opportunity to develop innovative custom software solutions. You will be instrumental in designing, coding, and enhancing components across various systems and applications. Utilizing modern frameworks and agile methodologies, your work will deliver scalable, high-performing solutions precisely tailored to unique business requirements.

This position requires a minimum of 2 years of experience. The ideal candidate will possess 15 years of full-time education and a strong command of Python programming.

Design, develop, test, and maintain cutting-edge Generative AI and Agentic AI solutions, leveraging Python and appropriate AI frameworks.

Develop sophisticated AI agents, intricate agentic workflows, and seamless integrations with LLMs, APIs, enterprise applications, databases, knowledge bases, and external tools.

Design and optimize prompts, implement advanced RAG and retrieval mechanisms, and actively experiment with models, tools, and agentic approaches to elevate solution quality.

Implement robust testing, validation, error handling, logging, and guardrails. Troubleshoot AI-generated outputs and collaborate effectively with architects, developers, QA teams, and business stakeholders.

Demonstrate strong proficiency in Python, with proven experience in developing production-quality applications, APIs, reusable components, and in testing, debugging, and performance optimization.

Possess hands-on experience in GenAI/LLM application development, including prompt engineering, context management, understanding tokens, model parameters, embeddings, and tool/function calling.

Acquire hands-on experience developing AI agents and agentic workflows using leading frameworks such as LangChain, LangGraph, AutoGen, CrewAI, Semantic Kernel, Amazon Bedrock Agents, or similar platforms.

Exhibit a good understanding of REST APIs, LLM integrations, RAG, embeddings, vector databases, semantic search, and knowledge retrieval principles.