Senior Engineer
Description
As a member of the AI team at Tejas Networks, you will work on the UFS (Unified Forecasting System) platform, building and maintaining AI-driven forecasting and anomaly detection solutions for network telemetry. The role involves developing REST API-based microservices for forecasting at Node-Port-KPI granularity and owning the complete lifecycle of services, from design and development to production support and platform integration.
You will work closely with observability platform and be responsible for ensuring reliable, scalable, and production-ready AI services.
Design, develop, and maintain Python-based microservices deployed using Docker and Kubernetes.
Contribute to existing Universal Forecasting Solution Platform by adding new features and use ecases.
Build and enhance REST API-based forecasting services for wireless and wireline network telemetry data.
Own services end-to-end, including architecture, implementation, deployment, monitoring, debugging, and production support.
Manage Kafka-based bulk forecasting pipelines, including:
Multi-topic architectures
Combination generation workflows
Consumer offset management
Token caching and performance optimization
Continuously improve data preprocessing services through research, experimentation, and implementation of new algorithms.
Integrate time-series foundation models into standalone, reusable forecasting microservices.
Design and implement large-scale anomaly detection systems across thousands of time series using statistical and machine learning techniques.
Review application code, deployment scripts, and infrastructure configurations.
Profile application performance, optimize resource utilization, and troubleshoot production bottlenecks.
Collaborate with platform teams to integrate AI services with systems such as Report Tool and TOP.
Must Have Skills
Strong programming experience in Python.
Applied mathematics, statistics, linear algebra, Supervised/unsupervised learning, model training, evaluation, and hyperparameter tuning.
Time‑series forecasting models: ARIMA, Prophet, LSTM, GRU, Transformer‑based models, handling missing values, normalization, feature selection, lag/rolling features, evaluation metrics like MAE, RMSE, MAPE, directional accuracy
Exposure to MLOps, model deployment, and monitoring practices.
Strong understanding of REST API design and microservice architecture.
Hands-on experience with Docker and Kubernetes.
Experience with CI/CD pipelines and version control systems (Git).
Strong troubleshooting and debugging skills in distributed systems.
Ability to independently investigate and resolve production issues.
Skills
Sign in to read and join the discussion.