Job Overview

Softobiz is hiring five AI Engineer Interns for a hands-on internship in Kochi focused on agentic AI, multi-agent systems, context engineering, and large language model integration. This full-time opportunity is intended for final-year students and recent graduates who want to build practical experience with production-oriented AI systems while working under the guidance of experienced engineers.

The internship provides exposure to how modern AI applications are designed beyond basic model usage. Interns will contribute to agent workflows, context management, retrieval, evaluation, verification, and LLM-powered features. The role emphasizes strong computer science fundamentals, data structures and algorithms, Python programming, analytical thinking, and the ability to learn quickly.

Key Responsibilities

• Assist senior engineers in designing and implementing multi-agent workflows using Python and LangGraph.
• Help create stateful workflows that support branching, looping, retries, checkpointing, and resumable execution.
• Contribute to safe pause and resume mechanisms and human-in-the-loop checkpoints.
• Learn and apply context engineering concepts such as layered context, retrieval, indexing, filtering, summarization, and context compaction.
• Help create token-aware prompts and typed context structures so individual agent steps receive relevant information.
• Support integration with LLM providers such as Anthropic, OpenAI, and Azure OpenAI.
• Work with prompt engineering, tool calling, structured output, and model-routing experiments.
• Assist with vector search, embeddings, and retrieval systems used to work with large codebases and other information sources.
• Help build evaluation and error-analysis processes to identify failure patterns and improve AI reliability.
• Contribute to verification, validation, deterministic output checks, observability, auditability, and reproducibility.
• Collaborate with platform and infrastructure engineers on deployment, inference, and persistence-related tasks.
• Participate in design reviews, code reviews, knowledge sharing, and internal demonstrations.
• Document experiments, including unsuccessful approaches, so the team can learn and improve future implementations.

Required Skills

Strong computer science fundamentals are essential for this internship. Candidates should understand data structures, algorithms, complexity analysis, and structured problem solving.

Python 3.10+ is the primary programming requirement. Candidates should be comfortable writing clean, readable, idiomatic Python and should have familiarity with concepts such as asynchronous programming and typing.

Hands-on exposure to LLMs, agentic AI, or AI development through academic projects, personal projects, coursework, or self-learning is valuable. Candidates should demonstrate genuine interest in learning multi-agent architectures and context engineering.

Preferred Skills

Experience with LangGraph, state machines, checkpointers, or human-in-the-loop workflows is beneficial. Knowledge of context selectors, filters, summarization, compaction, token budgeting, tool calling, structured outputs, and verification patterns can strengthen an application.

Exposure to Anthropic, OpenAI, or Azure OpenAI SDKs and prompt engineering is preferred. Familiarity with Pydantic v2, JSON Schema, typed contracts, vector databases, embeddings, and retrieval systems is also useful.

Candidates may benefit from experience with Qdrant, Azure AI Search, Model Context Protocol, CrewAI, Microsoft Agent Framework, or Temporal. Awareness of vLLM concepts such as batching, paged attention, quantisation, inference, and model routing is another advantage.

Education

Candidates should be pursuing or have recently completed a B.Tech, B.E., M.Tech, or MCA in Computer Science or a related discipline, or possess an equivalent qualification. The role welcomes final-year students and recent graduates.

Strong academic fundamentals are valued, while competitive programming experience through platforms or activities such as Codeforces, LeetCode, ICPC, or similar competitions is a strong plus.

Experience

This is an internship opportunity with an indicated experience range of 0 to 1 year. Professional experience is not the primary requirement. The focus is on strong fundamentals, programming ability, problem-solving skills, curiosity, and practical exposure to AI or software development.

Required Technologies

• Python 3.10+
• LangGraph
• Pydantic v2
• JSON Schema
• Anthropic
• OpenAI
• Azure OpenAI
• LLMs
• Prompt engineering
• Tool calling
• Structured output
• Vector search
• Embeddings
• Qdrant
• Azure AI Search
• Model Context Protocol (MCP)
• CrewAI
• Microsoft Agent Framework
• Temporal
• vLLM
• Azure
• AWS
• GCP

Soft Skills

The role requires a structured and analytical approach to software design, debugging, experimentation, and root-cause analysis. Clear written and verbal communication is important because interns will participate in reviews, demonstrations, documentation, and collaboration with technical teams.

Coachability and willingness to learn are especially important. Candidates should be comfortable receiving feedback, improving their work, sharing knowledge, and contributing to team standards. A collaborative mindset and genuine curiosity about emerging AI technologies will help interns succeed.

Benefits of Working in this Role

This internship offers practical exposure to production-oriented agentic AI development rather than limiting learning to academic exercises. Interns can work alongside experienced engineers and gain experience in multi-agent orchestration, LLM integration, retrieval, context engineering, evaluation, observability, and AI reliability.

The role can also help candidates develop professional software engineering habits through code reviews, design discussions, documentation, experimentation, and collaboration with infrastructure teams. Strong performance may provide a foundation for pursuing an AI Engineer or related software engineering career, although conversion to a full-time position should not be assumed.

Work Mode

The supplied job description does not specify whether the internship is onsite, hybrid, or remote. For SoftoJobs classification, the role is recorded as onsite based on the available information. Candidates should confirm the current work arrangement with Softobiz during the hiring process.

Location

The position is listed with Softobiz in Kochi, Kerala, India.

Who Should Apply

This opportunity is suitable for final-year Computer Science students, recent graduates, and early-career candidates who have strong DSA and Python fundamentals and want to specialize in AI engineering.

Applicants should be able to demonstrate programming projects, academic work, AI experiments, or self-learning that show practical problem-solving ability. Exposure to LLMs, agentic AI, LangGraph, retrieval, or related technologies is useful but should be represented accurately.

Career Growth

The internship can provide a foundation for careers in AI engineering, machine learning engineering, software engineering, LLM application development, agentic AI development, data and retrieval engineering, or AI platform engineering. Experience with production-oriented workflows, evaluation, reliability, and modern AI infrastructure can help interns build skills that are relevant to emerging AI engineering roles.

Application Advice

Keep your resume focused on evidence of strong fundamentals and practical work. Highlight Python projects, data structures and algorithms experience, academic projects, competitive programming achievements, and any LLM or agentic AI experiments.

If you have used LangGraph, OpenAI, Azure OpenAI, Anthropic, Pydantic, vector databases, MCP, Temporal, CrewAI, or related technologies, describe what you actually built rather than listing the tools alone. Include links to relevant projects or portfolios when appropriate.

Candidates with AI/ML or agentic-AI certifications from recognized learning providers may mention them. Cloud fundamentals certifications covering Azure, AWS, or GCP can also be included if genuinely completed.