AI Engineer Intern

About the role

Location: Remote | Hours: 20–40/week | Rate: $30/hour | Growth & Equity Potential About Incepta Incepta is an AI implementation firm that builds custom solutions for independent insurance agencies and launches products when we spot high-impact opportunities. We turn real-world business inefficiencies into AI agents that deliver real value. We’re not looking for just another intern—and you’re not here for busywork. We want prodigies: builders who’ve shipped side projects, won hackathons, or built AI agents from scratch. If you’ve built ambitious projects, shipped code that matters, or taught yourself to build AI agents, you’ll fit right in. You know how to move aggressively, learn fast, and take responsibility. Here, you’ll build agents for real customers—not just shadow someone else’s work. This isn't a run-of-the-mill internship; it's a launchpad for future technical leaders. You'll architect end-to-end AI agents, work alongside cracked engineers, and drive solutions from concept to deployment. Stand out, and you'll earn a permanent role with real equity as we scale together. At Incepta, we don't just experiment with AI—we automate complex workflows, replace manual processes, and build scalable, production-ready solutions. If you want to build something real, thrive on challenging technical problems, and solve cutting-edge AI development challenges, we want to hear from you. Role & Responsibilities • Design & ship AI agents end‑to‑end—from spec to production deployment • Orchestrate LLM pipelines (GPT‑4o, o1-preview, Claude 3.5 Sonnet, Gemini 2.0) with LangChain/LangGraph and vector search • Integrate APIs (OpenAI, Anthropic, custom services) with proper auth, error handling, and logging • Build lightweight UIs with React + Tailwind enabling users to trigger and monitor agents • Deploy & scale on serverless platforms; optimize for speed, accuracy, and cost • Implement function calling & tool use for agent-environment interactions via Model Context Protocol (MCP) • Build computer vision pipelines for document processing (PDF parsing, OCR, structured data extraction) • Monitor & iterate with performance tracking and weekly improvements Required Skills (Applicants must meet ALL criteria to be considered) • Enrollment: Currently pursuing Master's or PhD in Computer Science, AI, or related technical field • Programming: Proficient in Python or JavaScript (Node.js) • Prompt Engineering: Near-native English proficiency; get models to follow complex instructions and handle open-ended tasks • LLM APIs: Hands-on experience with OpenAI APIs (GPT-4 family, o1-preview, Function-Calling, Vision, Realtime API) • Multimodal AI: Experience with additional LLM providers (Anthropic Claude, Google Gemini) • RAG Implementation: Hands-on experience with retrieval-augmented generation, vector search, and document processing pipelines • Function Calling & Tool Use: Built agents that interact with external APIs, databases, and services • Serverless Deployment: Deploy code on AWS Lambda, GCP Functions, Azure Functions, or Firebase • Data Systems: Experience with SQL/NoSQL databases and vector databases (Pinecone, ChromaDB, FAISS) • Full-Stack Integration: Connect backend AI logic with frontend interfaces • Agent Frameworks: Experience with LangChain, LangGraph, LlamaIndex, or similar orchestration tools Bonus Skills (Nice to have, not required) • Computer Vision Integration: Combining vision models with LLMs for document/image analysis • Model Context Protocol (MCP): Structured agent-environment interaction protocols • Real-time AI Systems: Streaming, WebSocket, or low-latency AI applications • Containerization: Docker and basic Kubernetes • React/Vue: Building dashboards and chat interfaces • SaaS Scaling: Taking prototypes to production-ready systems • AI Safety: Responsible AI principles and data privacy practices SkillsPython · Node.js · TypeScript · REST/GraphQL APIs · LLM APIs (GPT-4o, o1-preview, Claude 3.5, Gemini) · Function Calling & Tool Use · RAG Implementation · LangChain/LangGraph/LlamaIndex · Model Context Protocol (MCP) · Computer Vision/Document Processing · AWS Lambda · GCP Cloud Functions · Azure Functions · Full‑Stack (React + Vite) · SQL & NoSQL · Vector Databases (Pinecone/ChromaDB/FAISS) · Multi-agent Orchestration · Docker · CI/CD · Prompt Engineering Originally posted on Himalayas