San Francisco, CA · available for new opportunities

Isha Valiveti

AI/ML Engineer

AI/ML Engineer with 5 years of experience building and scaling machine learning, generative AI, and LLM-powered solutions across cloud-native environments. I've shipped production-grade LLMOps platforms, RAG systems, and distributed data pipelines that improved model accuracy, cut inference costs, and held up under real user load — using Python, Databricks, MLflow, LangChain, AWS, and Apache Spark.

View Projects Download Resume

01 / selected work

Projects

Work outside the day job — experimentation, computer vision, and data storytelling.

LLM Energy Optimisation (Green LLM)

Structured experimentation framework built on IBM watsonx to evaluate prompt performance across multiple LLM configurations, with reproducible benchmarking workflows for prompt engineering and evaluation consistency.

IBM watsonxPrompt EngineeringLLMOps

Automated experiment tracking for data-driven prompt & model selection

Precision Agriculture: UAV-Based Computer Vision

Trained and fine-tuned a YOLOv5 detector to count mangoes in drone imagery for yield estimation, with a full pipeline from raw UAV frames to crop counts and on-device detection for live field estimates.

YOLOv5Computer VisionEdge Inference

0.91 mAP@0.5 on held-out fields

Netflix Content Strategy Trend Analysis

Analyzed 7,800+ titles with EDA, regression, and ARIMA to track the catalog's shift from film to TV, then built Tableau and Python dashboards that made international-expansion and content-mix trends readable for non-analysts.

ARIMARegressionTableau

Projected movie share falling from 36% to ~12% by 2031

02 / capabilities

Skills

The stack behind production RAG and agentic systems — from data engineering to the observability that keeps them honest.

Programming & Software Engineering

PythonJavaSQL PySparkSpring BootFastAPI REST APIsGitLinux

ML & Deep Learning

Time Series ForecastingAnomaly Detection Predictive MaintenanceLSTMXGBoost TensorFlowPyTorchScikit-learn Hyperparameter Tuning

Generative AI & LLMs

RAGAI AgentsLangChain LangGraphPrompt EngineeringPrompt Evaluation OpenAI APIsVector SearchSemantic Search Function / Tool CallingHybrid SearchReranking

MLOps & LLMOps

MLflowExperiment TrackingPrompt Versioning Model MonitoringModel GovernanceHuman Feedback Loops CI/CDAirflowA/B TestingModel Registry

Data Engineering & Big Data

Apache SparkDelta LakeDatabricks SQL Apache KafkaSpark StreamingETL Pipelines Feature EngineeringData Quality Validation

Cloud, DevOps & Observability

AWS (S3, ECS, Lambda, CloudWatch)DockerKubernetes OpenTelemetryDistributed TracingPostgreSQL RedisVector Databases

03 / career

Experience

Feb 2025 – Present · San Francisco, CA

AI/ML Engineer Databricks

Build LLMOps workflows and end-to-end GenAI lifecycle tooling — experiment tracking, prompt versioning, evaluation, and deployment monitoring — for AI applications used by over 120,000 users.

Designed agent evaluation frameworks integrating LangGraph, RAG pipelines, and observability metrics, improving latency visibility and tool-call reliability.
Built scalable data processing workflows with Apache Spark and Databricks SQL to prepare enterprise knowledge sources for retrieval-augmented generation.
Architected Spring Boot microservices and REST APIs for AI orchestration across model-serving endpoints and vector databases on AWS.
Deployed cloud-native workloads on AWS ECS, Lambda, S3, and CloudWatch, with Docker, Kubernetes, Kafka, and OpenTelemetry for observability and fault tolerance.

+31% response accuracy -28% inference cost -42% response latency 120K+ users served

Aug 2021 – Jul 2024 · India

Machine Learning Engineer Accenture

Built forecasting, predictive maintenance, and real-time anomaly detection systems on telecom network telemetry, supporting over 25 million subscribers.

Developed LSTM and XGBoost traffic forecasting models and AI-driven predictive maintenance pipelines over telecom infrastructure telemetry.
Implemented real-time anomaly detection with Python, Kafka, and Spark Streaming across billions of daily network events.
Built GenAI-powered operational assistants with LangChain, vector databases, and RAG to speed up troubleshooting.
Ran end-to-end MLOps pipelines on AWS using Airflow, MLflow, Docker, and Kubernetes, cutting deployment cycles by 60%.

+22% forecast accuracy -35% equipment failures -48% detection latency -60% deployment cycle time 25M+ subscribers supported

04 / one-pager

Resume

Want the full one-page version?

A condensed PDF summary of experience, skills, and outcomes — ready to share with recruiters.

// last updated: July 2026

Download Resume ↓

Education

Master of Professional Studies (MPS) in Data Science

University at Buffalo, The State University of New York — Buffalo, NY

Certifications

Databricks Certified Generative AI Engineer Associate
AWS Certified Machine Learning Engineer – Associate
AWS Generative AI Engineer Certification
Certified Kubernetes Administrator (CKA)