Job Overview
Role: AI Engineer Intern Location: Kochi Experience: 0-1 Year (Freshers) Qualification: B.Tech/B.E./M.Tech/MCA in Computer Science or related field Key Skills: Python, DSA, AI Engineering, LLM Integration, Agent Orchestration
Job Description
Softobiz is hiring five AI Engineering Interns for its Kochi team, open to final-year students and recent graduates pursuing or having recently completed a B.Tech/B.E./M.Tech/MCA in Computer Science or a related field. This is a hands-on, mentored internship focused on building production-grade agentic AI systems. Interns will work alongside senior engineers on multi-agent orchestration, context engineering, LLM integration, retrieval, evaluation, and AI-system reliability. The role places particular importance on computer-science fundamentals, data structures and algorithms (DSA), Python programming, problem-solving, and the ability to build working projects.
Roles and Responsibilities
- Agent Orchestration and Workflow Engineering: Assist with designing and implementing multi-agent workflows using LangGraph and Python, modeling AI processes as stateful graphs with multiple steps, transitions, branching, loops, retries, checkpointing, resumable execution, and human-in-the-loop checkpoints.
- Context Engineering: Learn to manage information supplied to an AI model at each stage of an agent workflow, including layered context, context selection, filtering, retrieval and indexing, active working sets, token budgeting, summarization, history compaction, and typed context schemas.
- LLM Integration: Contribute to integrating LLM providers such as Anthropic, OpenAI, and Azure OpenAI, working with prompt engineering, tool calling, structured output, provider SDKs, and model routing.
- Retrieval and Embeddings: Build retrieval capabilities using vector search and embeddings (e.g., Qdrant, Azure AI Search), understanding the basic RAG flow from documents to LLM context, and code-intelligence techniques.
- Model Routing: Contribute to model-routing experiments, selecting different models based on factors such as task type, latency, cost, and quality requirements to balance performance and economics.
- Agent Evaluation and Error Analysis: Build evaluation and error-analysis loops for AI systems, treating failures as useful feedback to systematically improve reliability.
- Verification and Validation: Implement verification and validation patterns and deterministic gates for agent outputs to reduce the risk of accepting incorrect or unsafe model responses, using schema validation, rule-based checks, and output constraints.
- Observability and Governance: Emphasize making agent decisions and context observable, auditable, and reproducible, which is crucial for understanding agent behavior in production AI systems.
Skills and Eligibility Criteria
Educational Background: Pursuing or recently completed B.Tech/B.E./M.Tech/MCA in Computer Science or a related field, or equivalent. Final-year students and recent graduates are welcome.
Experience: 0–1 Year. Exposure to LLM or agentic AI projects through academics or self-learning is beneficial.
Mandatory Technical Skills:
- Strong data structures, algorithms, complexity analysis, and problem-solving skills
- Hands-on experience with Python 3.10+ (async programming, typing, clean code)
- Strong computer-science fundamentals
Competencies:
- Strong willingness to learn LangGraph and context-engineering concepts
- Exposure to LangGraph, multi-agent design, tool calling, structured output, verification, Anthropic, OpenAI, Azure OpenAI SDKs, prompt engineering, Pydantic v2, JSON Schema, typed contracts, vector stores, Qdrant, Azure AI Search, embeddings, Model Context Protocol (MCP), CrewAI, Microsoft Agent Framework, Temporal, vLLM, paged attention, batching, quantization, model routing