I'm Anjaney Shrivastava—an engineer working at the intersection of dependable AI systems, software architecture, and experimental research. I like turning ambitious ideas into things that can be tested.
A home for published work and early-stage ideas. Clear distinction between what's public and what's still being explored.
A Task-Adaptive, Code-Based Ensemble for Proactive Verification of LLM Outputs. Exploring a practical direction for verifying language-model responses before they are trusted downstream.
An exploratory research direction in affective audio understanding, asking what human humming might communicate. Research details are intentionally limited while the work develops.
My work centers on reliable pipelines, orchestration systems, and the engineering decisions behind real-world AI applications.
Designing modular workflows that connect preprocessing, extraction, verification, classification, and downstream actions.
Developing systems with reusable components, controllable execution paths, and structured observability.
Optimizing execution patterns, handling failures, and making production behavior more measurable and predictable.
I care about ideas that survive contact with reality.
My background spans electronics and communication engineering and enterprise AI development. Today, I'm interested in verification, adaptive systems, applied research, and the space between a promising concept and a dependable implementation.
This website will evolve into an open notebook of research updates, selected projects, and technical writing.
Interested in research collaboration, rigorous experiments, or engineering opportunities? My first public paper is a good place to start a conversation.