Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

AI/ML Engineer with 4+ years building production-grade AI systems across GCP-native platforms. Specialized in multi-agent architectures (Google ADK, A2A, MCP), LLM-powered RAG pipelines, and real-time anomaly detection using LSTM/Transformer models reducing false-positive alerts by 35% and improving Mean Time to Detect by 20% at Datadog. Proven track record in full-stack ML systems, from Kafka-based streaming pipelines to FastAPI/React predictive maintenance platforms processing 50M+ IoT records, cutting unplanned downtime by 23%. Skilled in CI/CD, Vertex AI, and cross-cloud agent orchestration for enterprise-scale reliability…AI/ML Engineer with 4+ years building production-grade AI systems across GCP-native platforms. Specialized in multi-agent architectures (Google ADK, A2A, MCP), LLM-powered RAG pipelines, and real-time anomaly detection using LSTM/Transformer models reducing false-positive alerts by 35% and improving Mean Time to Detect by 20% at Datadog. Proven track record in full-stack ML systems, from Kafka-based streaming pipelines to FastAPI/React predictive maintenance platforms processing 50M+ IoT records, cutting unplanned downtime by 23%. Skilled in CI/CD, Vertex AI, and cross-cloud agent orchestration for enterprise-scale reliability

Ishaan Bandekar

AI Engineer, Data Scientist, Full Stack Developer, +3





AI/ML Engineer with 4+ years building production-grade AI systems across GCP-native platforms. Specialized in multi-agent architectures (Google ADK, A2A, MCP), LLM-powered RAG pipelines, and real-time anomaly detection using LSTM/Transformer models reducing false-positive alerts by 35% and improving Mean Time to Detect by 20% at Datadog. Proven track record in full-stack ML systems, from Kafka-based streaming pipelines to FastAPI/React predictive maintenance platforms processing 50M+ IoT records, cutting unplanned downtime by 23%. Skilled in CI/CD, Vertex AI, and cross-cloud agent orchestration for enterprise-scale reliability…AI/ML Engineer with 4+ years building production-grade AI systems across GCP-native platforms. Specialized in multi-agent architectures (Google ADK, A2A, MCP), LLM-powered RAG pipelines, and real-time anomaly detection using LSTM/Transformer models reducing false-positive alerts by 35% and improving Mean Time to Detect by 20% at Datadog. Proven track record in full-stack ML systems, from Kafka-based streaming pipelines to FastAPI/React predictive maintenance platforms processing 50M+ IoT records, cutting unplanned downtime by 23%. Skilled in CI/CD, Vertex AI, and cross-cloud agent orchestration for enterprise-scale reliability

Available to hire

AI/ML Engineer with 4+ years building production-grade AI systems across GCP-native platforms. Specialized in multi-agent
architectures (Google ADK, A2A, MCP), LLM-powered RAG pipelines, and real-time anomaly detection using LSTM/Transformer models reducing false-positive alerts by 35% and improving Mean Time to Detect by 20% at Datadog. Proven track
record in full-stack ML systems, from Kafka-based streaming pipelines to FastAPI/React predictive maintenance platforms
processing 50M+ IoT records, cutting unplanned downtime by 23%. Skilled in CI/CD, Vertex AI, and cross-cloud agent orchestration for enterprise-scale reliability

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

Work Experience

AI/ML Engineer at Datadog USA

July 1, 2025 - Present

Architected a real-time anomaly detection engine combining STL decomposition with LSTM and Transformer forecasting models to detect infrastructure degradation, reducing false-positive alerts by 35% in multi-tenant production environments. Designed and deployed a multi-agent AI system on GCP using Google ADK to enable incident investigation and management conversationally. Built an internal SDK/library to standardize agent creation, configuration, and tool integration, and implemented A2A orchestration so specialized agents collaborate across domains during live incidents. Implemented Model Context Protocol (MCP) enabling secure agent calls to external systems and APIs, and built a RAG pipeline over incident reports, runbooks, and logs for grounded agent responses. Established OpenTelemetry-based observability and BigQuery cost monitoring for LLM usage, latency, and performance. Engineered Kafka-based streaming pipelines for ingestion, feature computation, inference, and agent-triggered

Machine Learning Engineer at Infosys

March 1, 2020 - July 1, 2023

Designed and developed a full-stack predictive maintenance platform with FastAPI APIs and a React dashboard integrating XGBoost, Random Forest, and LSTM models for failure classification and Remaining Useful Life (RUL) forecasting. Built Python ETL pipelines processing 50M+ IoT sensor records, including rolling statistics, lag features, and FFT-based frequency-domain indicators via containerized feature engineering jobs. Trained, tuned, and evaluated multiple models using cross-validation and hyperparameter optimization, selecting production models using precision/recall trade-offs. Implemented anomaly detection over telemetry to flag abnormal equipment behavior, contributing to a 30% reduction in emergency response lead times and a 23% reduction in unplanned downtime. Built and maintained backend services serving predictions with validation and authentication, integrating with SQL Server/PostgreSQL. Developed frontend UI components for equipment health dashboards and real-time visuali