Python AI Engineer Foresight

Bengaluru - India

Monthly Salary: Not Disclosed

Posted on: Yesterday

Vacancies: 1 Vacancy

Job Summary

Key Responsibilities

End-to-End Model Training: Design train and fine-tune Machine Learning and Deep Learning models from scratch. You know how to select the right architecture (CNN RNN Transformer) for the problem.
Inference & Deployment: You dont just stop at training. You are responsible for running models in production optimizing them for latency and exposing them via APIs (FastAPI/Flask).
Deep Understanding: You can debug a model not just by changing code but by analyzing loss curves adjusting learning rates and fixing data imbalances. You know why the model is failing.
Code Quality: Write clean modular and production-ready Python code. Your code is testable version-controlled and scalable.
Learn & Adapt: Collaborate with seniors to learn Knowledge Graph technologies (Neo4j RDF) and apply your AI skills to graph-based problems (e.g. Graph Neural Networks).

Must-Have Skills

Fundamental knowledge or AWS or a similar cloud platform
AI & Math Fundamentals: Strong grasp of the theory behind MLyou understand gradient descent backpropagation activation functions and overfitting/underfitting concepts.
Deep Learning Frameworks: 3 years of experience with PyTorch or TensorFlow. You can write custom training loops and data loaders.
Python Mastery: Expert-level Python skills. You understand object-oriented programming decorators and memory management.
Data Engineering for AI: Ability to build efficient data pipelines (Pandas/NumPy) to preprocess complex datasets before feeding them into models.
Model Evaluation: Experience setting up robust validation strategies (Cross-validation F1-score AUC-ROC) to ensure models actually work on unseen data.

Nice to Have

Experience with LLMs (Large Language Models) or NLP.
Exposure to graph databases (Neo4j Neptune) or network analysis.
Experience deploying models using Docker or Kubernetes.

Required Experience:

Manager

Key ResponsibilitiesEnd-to-End Model Training: Design train and fine-tune Machine Learning and Deep Learning models from scratch. You know how to select the right architecture (CNN RNN Transformer) for the problem.Inference & Deployment: You dont just stop at training. You are responsible for runnin...