Open to opportunities

Anish Shivamurthy

@anishshivamurthy

Message

AI systems engineer building reliable GraphRAG LLM platforms with full-stack observability.

United States

Message

What I'm looking for

I want to build reliable, evaluation-driven AI products—shipping backend APIs plus full-stack observability for retrieval, RAG/GraphRAG, and multi-provider reliability—on a team that values deterministic testing and clean, scalable data pipelines.

I’m an AI systems engineer and software developer focused on turning research-grade ideas into production-style reliability. I’m currently pursuing an M.S. in Computer Science at the University of California, Irvine, with coursework spanning algorithms, AI/ML, and transactions processing and data management.

In my recent roles, I strengthened both engineering fundamentals and evaluation discipline—mentoring 200+ students through C++, Java, and Python while running a Linux/Bash auto-grading system with strict flags and deterministic edge-case validation. I also built TypeScript/Node.js REST APIs and Python ETL workflows using PostgreSQL staging tables to improve data cleanliness and reduce backend friction.

My work increasingly centers on trustworthy AI: I architected a self-healing, graph-augmented LLM reliability mesh with adaptive retrieval, deterministic LLM-as-a-judge evaluation, semantic retry orchestration, and multi-provider failover, backed by GraphRAG (Neo4j/Cypher) and full-stack Next.js/TypeScript observability. I’m also developing scalable collaborative AI/ML workflow capabilities in an open-source platform, integrating backend Scala descriptors with Python-based visualization (Plotly) and end-to-end schema validation for live UI synchronization.

Experience

Work history, roles, and key accomplishments

Current

AI Systems Engineer

Current

Aegis

May 2026 - Present (1 month)

Architecting a production-style AI reliability platform providing adaptive retrieval, deterministic LLM evaluation across multiple quality dimensions, semantic retry orchestration, and health-aware multi-provider failover. Implemented a hybrid GraphRAG retrieval engine with Neo4j multi-hop traversal and FAISS search, validated through 190+ deterministic offline tests.

Python fastAPI Faiss Sentence Transformers Neo4j Graph RAG LLM Evaluation Prometheus

Current

AI/ML Engineer

Current

Texera.io

Jan 2026 - Present (5 months)

Developing features in Apache-incubating Texera for building and sharing collaborative AI/ML visual workflows at scale, including a custom scientific visualization Polar Chart operator. Implemented full-stack workflow support for importing external datasets (Hugging Face, Kaggle) with schema validation, metadata parsing, and real-time UI synchronization across Scala and TypeScript layers.

Scala Python Plotly Apache Texera TypeScript Data Visualization

Instructional Assistant

UCSC Baskin School of Engineering

Jan 2024 - Jun 2025 (1 year 5 months)

Mentored 200+ students in C++, Java, and Python across Data Structures, Artificial Intelligence, and Computer Architecture, strengthening understanding of algorithms, memory management, and debugging. Operated and extended a Linux/Bash auto-grading system to deterministically evaluate visible/hidden tests with strict compilation flags and timeouts.

C#Java Python BASH Auto Grading Systems Data Structures Algorithms Debugging Memory Management

Software Engineer Intern

Tech4Good Lab, UCSC

Jan 2024 - Jun 2024 (5 months)

Developed TypeScript/Node.js REST APIs for onboarding and team-assignment workflows, adding request validation, consistent error handling, and improved database queries. Built Python ETL workflows backed by PostgreSQL staging tables to validate multi-source partner data for a 150+ student lab and prevent recurring dashboard breakages.

TypeScript Node REST APIs Error Handling PostgreSQL ETL Data Modeling API Integration Agile

Computer Vision Researcher

3D Faces Lab, UCSC

Mar 2023 - Jun 2023 (3 months)

Prototyped a browser-based annotation interface for facial attributes to standardize labeling across a 2,000+ image dataset using validation constraints and structured input flows. Refactored preprocessing and storage pipelines to improve dataset consistency and memory efficiency, contributing to a 46% downstream ResNet-152 performance gain for avatar attribute generation.

Computer Vision Browser Based Annotation UI Data Validation data preprocessing ResNet 152 Python

Education

Degrees, certifications, and relevant coursework

University of California, Irvine

Master of Computer Science, Computer Science

Expected to complete a Master of Computer Science at the University of California, Irvine, with coursework including Algorithms, Artificial Intelligence, Machine Learning, and Advanced Problem Solving and Programming.

University of California, Santa Cruz

Bachelor of Science, Computer Science

2021 - 2025

Completed a B.S. in Computer Science at the University of California, Santa Cruz from September 2021 to March 2025, covering Analysis of Algorithms, Data Structures, Computer Networks, Artificial Intelligence, Natural Language Processing, and Computer Systems.