Lanke Kiran Teja

Remote

Built AI-powered Flutter app for women's health tracking with support for 32+ Indian languages targeting rural and underserved communities.

• ·Built AI-powered Flutter app for women’s health tracking with support for 32+ Indian languages targeting rural and underserved communities.
• ·Integrated offline-first architecture and IoT medical device connectivity for reliable access in low-connectivity regions.
• ·Implemented domain-specific ML and LLM models for clinical report processing with 94% accuracy.

Hyderabad, India

Developed Challenges module across CMS and React Native including course linking, schedule management, and admin workflows.

• ·Developed Challenges module across CMS and React Native including course linking, schedule management, and admin workflows.
• ·Fixed school-level data isolation by implementing organization templates, enabling multi-tenant platform stability.
• ·Optimized 50+ Xano REST APIs, reducing response latency by 45% through indexing and query restructuring.
• ·Strengthened backend security with token authentication, input validation, and role-based access control.

Hyderabad, India

Deployed production LLM inference pipelines handling 1,000+ requests/minute with automated scaling and fallback routing.

• ·Deployed production LLM inference pipelines handling 1,000+ requests/minute with automated scaling and fallback routing.
• ·Optimized inference latency by 40% through KV-cache implementation and attention optimization under concurrent load.
• ·Built monitoring dashboards and health checks maintaining 99.5% uptime across deployments.

Remote

Developed ML classification models achieving 85–92% accuracy through feature engineering and systematic hyperparameter tuning.

• ·Developed ML classification models achieving 85–92% accuracy through feature engineering and systematic hyperparameter tuning.
• ·Constructed end-to-end pipelines covering data preprocessing, feature extraction, model training, and performance analysis.

Lucknow, India

Identified and resolved network bottlenecks, improving system throughput by 35% via QoS configuration and traffic optimization.

• ·Identified and resolved network bottlenecks, improving system throughput by 35% via QoS configuration and traffic optimization.
• ·Automated network monitoring and log analysis with Python scripts, reducing incident diagnosis time from hours to minutes.