AI Engineer

Location: Bengaluru, India

Department: Data Science

Experience: 2-4

• Tackle complex, open-ended problems where business processes must be reverse-engineered from noisy, incomplete, and high-volume event streams; there is no textbook answer -- you define the approach.
• Design efficient algorithms for process discovery at scale, handling billions of events, millions of unique traces, and thousands of activity variants while maintaining sub-second query performance.
• Design and implement models for mining, clustering, and classifying sequential activity data (event logs, user interaction sequences, clickstreams) at enterprise scale.
• Build ML models to classify and label user activities from raw desktop observation data (screenshots, UI metadata, application events) using computer vision and NLP.
• Train, evaluate, and deploy models including sequence-to-sequence architectures (Transformers, LSTMs, HMMs), graph neural networks and clustering algorithms for process pattern recognition.
• Apply techniques from temporal data mining, e.g., sequential pattern mining, episode mining, and time-series analysis to extract insights from event streams.
• Leverage LLMs and generative AI for intelligent summarization, anomaly explanation, and natural-language querying of process data.
• Debug and reason through edge cases in complex data pipelines where subtle algorithmic errors can cascade into incorrect process maps affecting enterprise decisions.
• Partner with product, UX, and customer-facing teams to translate business process challenges into well-defined algorithmic and ML problem statements.
• Contribute to Skan's agentic AI capabilities, building models that not only discover inefficiencies but recommend and orchestrate automated actions.
• Stay current with advances in AI. Research and experiment with state-of-the-art AI techniques to improve existing models and integrate cutting-edge techniques into production systems.

• Bachelor's/Master’s degree in Computer Science, Artificial Intelligence, Machine Learning, Data Science, or a related quantitative field.
• Exceptional problem-solving ability: You excel at breaking down ambiguous, complex problems into tractable sub-problems and designing clean algorithmic solutions.
• Strong proficiency in Python and experience with PyTorch or TensorFlow, scikit-learn, pandas, and SQL.
• Deep understanding of data structures, algorithms, and optimization techniques.
• Experience with data engineering, feature extraction, and model evaluation.
• Knowledge of deep learning techniques (CNNs, RNNs, Transformers, GANs, etc.).
• Familiarity with graph algorithms and graph-based modeling (traversal, shortest path, community detection, graph neural networks).
• Experience with LLMs and knowledge of multi-modal AI (text, image).
• Hands-on experience building and deploying ML/AI models or algorithm-heavy systems in production.

• Strong performance in competitive programming, algorithmic challenges, or systems design interviews is a signal we value.
• Familiarity with process mining concepts, event logs, process discovery algorithms, conformance checking, and variant analysis. Experience with PM4Py, ProM, or similar tools is a plus.
• Knowledge of cloud computing (AWS, GCP, Azure) and MLOps tools.
• Familiarity with agentic AI frameworks and LLM-based tool-use patterns (LangChain, function calling, RAG pipelines).