ML Engineer Intern

Aqyron Labs Private Limited; Kolkata District; On-site; INR 167,000 - 279,000; Part time

Job summary

Aqyron Labs Private Limited invites talented ML Engineer Interns to join our AI team in Kolkata. You will help build multi-agent LLM pipelines, curate domain datasets, and benchmark model performance. You’ll also contribute to deploying ML inference endpoints with Docker and FastAPI in production-grade environments.

Ideal candidates are pursuing degrees in CS/AI and are eager to ship real-world AI applications, with hands-on project experience in NLP, LLMs, and vector databases.

Qualifications

• Strong foundational knowledge in Python, data structures, and algorithms.
• Familiarity with ML/DL frameworks (PyTorch, TensorFlow, or Scikit-learn).
• Understanding of NLP, LLMs, embeddings, and vector databases.
• Hands-on project experience with APIs, Git version control, and Linux environments.
• Enthusiasm for shipping real-world AI applications and eager to learn production-grade software engineering.

Responsibilities

• Assist in building and evaluating multi-agent LLM pipelines and autonomous workflows.
• Collect, clean, preprocess, and benchmark datasets for fine-tuning and RAG.
• Implement automated evaluation benchmarks for model accuracy, latency, and hallucinations.
• Collaborate with senior AI engineers to containerize and deploy ML inference endpoints using Docker & FastAPI.
• Participate in research sprints on transformer architectures, quantization, and vector retrieval.

Skills

Python

Data structures

Algorithms

ML frameworks

NLP/LLMs

APIs & Git

Linux

Tools

Docker

FastAPI

PostgreSQL (pgvector)

AWS/GCP

Job description

Join Aqyron Labs AI team as an ML Engineer Intern. Work on fine-tuning LLMs, building agentic workflows, data pipelines, and scalable model inference for production systems.

What you will do

• Assist in building and evaluating multi-agent LLM pipelines and autonomous workflows
• Collect, clean, preprocess, and benchmark datasets for domain-specific fine-tuning and retrieval-augmented generation (RAG)
• Implement automated evaluation benchmarks to quantify model accuracy, latency, and hallucination rates
• Collaborate with senior AI engineers to containerize and deploy ML inference endpoints using Docker and FastAPI
• Participate in research sprints exploring modern transformer architectures, quantization techniques, and vector retrieval

Required qualifications

• Strong foundational knowledge in Python, data structures, and algorithms
• Familiarity with Machine Learning / Deep Learning frameworks (PyTorch, TensorFlow, or Scikit-learn)
• Understanding of modern NLP, Large Language Models (LLMs), embeddings, and vector databases
• Hands-on project experience with APIs, Git version control, and Linux environments
• Enthusiasm for shipping real-world AI applications and eager to learn production-grade software engineering

Preferred qualifications

• Prior experience or projects involving LangChain, LlamaIndex, Hugging Face Transformers, or vLLM
• Basic familiarity with Docker, FastAPI, PostgreSQL (pgvector), or cloud services (AWS/GCP)
• Contributions to open-source AI projects or published academic papers/competitions (e.g. Kaggle)