AI Engineer Intern

About the internship

This work-from-home internship is for an AI engineer who will help build the core systems behind adaptive learning products. The role focuses on designing agent workflows, retrieval-augmented generation (RAG) pipelines, evaluation and observability, and the infrastructure needed to run reliable AI capabilities at scale.

What you'll work on

• Design and implement agent orchestration with LangChain so an LLM can plan actions, use tools, review results, and decide the next step.
• Select the most suitable orchestration approach for each use case, including single-agent loops, supervisor/subagent structures, or graph-based multi-step workflows.
• Create and support RAG pipelines powered by vector databases.
• Build reusable, versioned agent skills that are narrow in scope, testable, and callable by other components.
• Set up evaluation and observability workflows with Langfuse to trace runs, score outputs, and detect regressions.
• Develop memory systems for agents using Mem0 and Zep so information persists appropriately across sessions.
• Write and maintain Python APIs that expose agent capabilities to the broader product.
• Determine when workflows should route to human review versus automated LLM-as-judge scoring, and implement both paths.
• Add guardrails at the moment the agent takes action, including tool-call authorization and output validation.
• Troubleshoot multi-step and multi-agent issues by tracing executions from end to end.
• Balance model choice across cost, latency, and capability, and route tasks to the most appropriate model.
• Deploy and scale agent systems on GPU infrastructure while keeping capacity, cost, and reliability in check.
• Maintain a tidy, version-controlled GitHub codebase with clear commits and review discipline so the system stays maintainable as models, tools, and provider APIs evolve.

Eligibility

Applicants must be able to work remotely, be available to start between 7 July 2026 and 11 August 2026, commit to the full 6-month duration, and bring relevant skills and interest in the domain. Women who are looking to begin or restart their careers are also welcome to apply.

Requirements

• Hands-on experience with multi-provider model routing for cost and latency optimization.
• Awareness of security risks that are specific to LLM and agent-based systems.
• Experience conducting both offline pre-deployment evaluations and online production evaluations.
• Exposure to alternative orchestration tools such as CrewAI, LangGraph, or similar frameworks.
• Experience with vector databases or memory systems beyond a primary stack.
• Prior experience building a similar system from scratch in an early-stage startup or company setting.

About the company

Vega Visionary is an EdTech company focused on adaptive learning. It uses advanced artificial intelligence to create personalized learning journeys that adapt to each learner’s needs. The platform identifies skill gaps and automatically produces targeted content to help close them.

Stipend and perks

The internship offers a monthly stipend in the range of INR 13,000 to 17,000. The compensation structure includes fixed pay of INR 12,000 to 15,000 per month and incentive pay of INR 1,000 to 2,000 per month. Other benefits include a certificate.

Additional information

This internship is listed as work from home. The application deadline is 7 August 2026 at 23:59:59.