Himani Ajit Mali

About Himani Ajit Mali

I’m driven by a simple belief: ’.As an, I explore the frontier of intelligent systems:multi-agent ecosystems, autonomous workflows, advanced retrieval architectures, and production-ready AI that solves real problems at scale.At PhysicianX, I’ve engineered %, powered by self-healing distributed architectures and LLM-based decision-making helping modernize one of the most complex industries in the world.I’m fascinated by systems that.My work spans multi-agent architectures, high-precision GraphRAG pipelines, transformer fine-tuning, inference optimization, and intelligent workflows that push the limits of what models can understand. I’ve also explored, studying how agents learn through experience to navigate complex environments.What drives me most is the idea that. I’m not just building models—I’m exploring how intelligence forms, evolves, and interacts inside complex digital systems. And I aim to build AI that is not only powerful but meaningful.Every day, I try to combine curiosity with courage:to push boundaries, to question assumptions, and to imagine AI not just as a tool, but as a new kind of collaborator.🫱‍🫲 :To create AI that amplifies human potential—and to learn, share, and build that future together.If you believe the next decade belongs to bold ideas, collaborative intelligence, and human-centered innovation,’. ’.

Experience

1. Founding Ai Engineer(Agentic Systems)

   Zavvis

   Jan 2026 — Present · Miami, FL, US

   Design and build the agentic architecture powering Zavvis’ Financial Data Observability Engine,including multi-agent workflows and orchestration.• Develop LLM-driven pipelines that integrate retrieval, event triggers, and financial logic forcontinuous monitoring across core financial systems.• Create retrieval and context modules (vector stores, embeddings, structured context) to supportaccurate, journal-entry–level reasoning.• Implement anomaly and variance detection, combining rules-based methods with ML to surfacematerial deviations and financial risks.• Build root-cause tracing agents that link anomalies to the exact source transactions andgenerate clear diagnostic explanations.• Develop the war-room presentation engine, enabling agents to produce real-time, drillableexecutive reports.• Collaborate with backend, ETL, and data science teams to align agent workflows with theUnified Transaction Model and ledger reconstruction pipeline.• Optimize and evaluate AI performance, ensuring reliability, accuracy, and audit-ready outputs ina production finance environment.

Education

• SIES Graduate School Of Technology

  Artificial Intelligence and Machine Learning

• SIES Graduate School Of Technology

  Bachelor of Engineering - BE, Information Technology

• University at Buffalo

  Master's degree, Artificial Intelligence