AI Engineer Intern

About ShyftLabs

Since its inception in 2020, ShyftLabs has specialized in leveraging data to fuel growth for Fortune 500 companies through cutting-edge digital innovations. With a presence across Canada, the United States, and India, the company is rapidly expanding and seeks driven, inquisitive individuals passionate about learning and solving complex technological challenges.

About Continuum

Continuum is a sophisticated AI agent execution and management platform that empowers teams to build, deploy, and operate dependable agentic applications. Its capabilities include agent orchestration, multi-model routing, persistent memory, tool integrations, executable workflows, governance policies, guardrails, performance evaluations, and observation tools.

Role Overview

We are seeking an AI Engineer Intern who is eager to develop and refine production-grade AI agents and applications. The selected candidate will contribute directly to the enhancement of the Continuum platform while utilizing it to craft agentic solutions addressing real-world enterprise challenges.

Key Responsibilities

• Design and refine agent orchestration and coordinate workflows involving multiple agents.
• Develop enterprise-focused agentic applications leveraging the Continuum system.
• Collaborate with a variety of commercial and open-source language models such as OpenAI, Anthropic, Gemini, Llama, Qwen, and Mistral.
• Optimize intelligent model routing by considering factors including task difficulty, quality of output, response latency, and operational costs.
• Establish and enhance persistent memory and state management systems for long-duration agent workflows.
• Create functionality for tool invocation and integrate with APIs, databases, and other enterprise platforms.
• Work with MCP servers and implement function tools to enable agent interactions with external services.
• Engineer pipelines for context management and retrieval using vector and graph database technologies.
• Implement AI safety measures including security protocols, access control, data privacy safeguards, and policy enforcement guardrails.
• Develop evaluation frameworks to assess metrics like accuracy, factual grounding, hallucination rates, tool utilization, workflow success, latency, and cost efficiency.
• Enhance visibility and traceability of agent decisions, tool usage, and workflow executions.
• Improve prompt engineering, optimize model deployments, token usage, system response times, and infrastructure expenses.
• Build APIs, backend infrastructure, and prototype user interfaces for agentic application deployment.
• Produce clean, reusable, thoroughly tested, and well-documented Python code.
• Contribute actively to the open-source Continuum codebase including examples, documentation, and improving the developer experience.

Candidate Requirements

• Pursuing or recently completed a degree in Computer Science, AI, Data Science, Software Engineering, or related discipline.
• Proficient programming skills in Python.
• Solid understanding of machine learning concepts, natural language processing, and large language model fundamentals.
• Experience developing at least one application powered by LLMs or featuring agentic characteristics.
• Familiarity with prompt engineering techniques, embeddings, retrieval-augmented generation (RAG), tool invocation, and structured output formats.
• Working knowledge of APIs, version control (Git), database technologies, and standard software engineering workflows.
• Ability to investigate complex technical challenges, experiment with various approaches, and communicate outcomes clearly.
• Genuine enthusiasm for building dependable AI systems beyond basic demonstrations.

Preferred Qualifications

• Experience with agent frameworks like LangGraph, LangChain, LlamaIndex, AutoGen, or CrewAI.
• Knowledge of commercial and open-source models such as OpenAI, Anthropic, Gemini, AWS Bedrock.
• Experience working with vector databases such as Milvus, Pinecone, Weaviate, or Chroma.
• Familiarity with graph databases including Neo4j.
• Skills in PostgreSQL, Redis, containerization (Docker), orchestration (Kubernetes), or cloud infrastructure management.
• Understanding of AI monitoring and evaluation tools like Langfuse.
• Knowledge concerning multi-tenancy, identity management, authorization, and enterprise security protocols.
• Expertise in model routing, inference optimization, prompt compression, or cost reduction strategies.
• Track record of contributions to open-source AI projects, research initiatives, hackathons, or technically challenging personal projects.

Compensation

The position offers hourly remuneration ranging between 20 and 30 Canadian dollars.

Work Environment and Benefits

• Hybrid work flexibly structured with three days per week on-site at their Toronto office downtown.
• Opportunity to work from a central location in Toronto's business district.
• Access to extensive learning and professional development tools to bolster skill advancement.

Diversity and Inclusion

ShyftLabs fosters an inclusive and diverse workplace, encouraging applications from candidates of all ethnicities, religions, disabilities, gender identities, orientations, family statuses, ages, and nationalities. Accommodations during recruitment are available upon request.