Prachotan Reddy Bathi

System Architect at Arm
Platform Security. System Manageability.

Explore my work

Each animal introduces a topic.Select one to learn more.

System architecture, from hardware to software.About my background
01 / Background

Curious
by nature.

I’m a System Architect at Arm. I help define security architecture and develop manageability prototypes.

I work with standards bodies and open-source communities across hardware, firmware, and software. I’m also exploring AI safety and agent security.

I enjoy teaching and advocate for digital literacy to address bias and barriers linked to unequal access to technology.

Get in touch
Based in
Austin, Texas
Education
Computer ScienceMaster’s degree · University of Southern California
What I enjoy
Pickleball, Teaching, Cooking

Standards & industry

  • OCP
  • TCG
  • UEFI
  • NIST

Open-source projects

  • Arm Trusted Firmware
  • EDK2
  • Linux
  • OpenBMC
02 / Work & interestsSix areas of work & interest.

A different breed
of thinking.

Architecture and prototyping at Arm.
Research interests and digital literacy advocacy.

01Architecture

See the whole forest.

Arm System Architecture

Defining interfaces and responsibilities across Arm systems.

Read more
02Security architecture

Trust starts at the roots.

Platform Security

Roots of trust, workload isolation, and remote attestation.

Read more
03Prototyping

Build. Observe. Improve.

System Manageability

Prototyping secure management and fault handling.

Read more
04Founder · Linepine

Wisdom is better shared.

Teaching & Advocacy

Bringing technology closer to children through Linepine.

Read more
05Research interest

Actions need boundaries.

AI Safety & Security

Exploring how to constrain and evaluate AI agents.

Read more
06Research interest

Grounded in hardware.

Hardware-enforced AI Safety

Exploring hardware-enforced constraints on AI systems.

Read more
Open to technical discussion and collaboration.Discuss a topic

Work & interests

Architecture

Arm System Architecture

My work at Arm spans the boundaries between hardware, firmware, and software. I help define system security requirements and interfaces, working with standards organizations and open-source projects to connect architectural specifications with implementation.

  • System architecture
  • Standards
  • Open source

Questions & considerations

  • How do you ensure an off-the-shelf OS boots on an Arm machine?
  • Where should security responsibilities and trust boundaries sit between components?
  • How can specifications define behavior precisely enough for interoperable implementations?
Back to all topics

Security architecture

Platform Security

I help define security architecture for Arm systems, including Trusted Platform Modules (TPMs), Caliptra, platform Root of Trust (pRoT), Arm Confidential Compute Architecture (CCA), and remote attestation. The architectural questions concern trust boundaries, workload isolation, and the evidence a platform provides about its state.

  • TPM
  • Caliptra
  • xRoT
  • Arm CCA
  • Remote attestation

Questions & considerations

  • What assumptions underpin the root of trust, and which components remain trusted?
  • What does attestation evidence establish, and how should a verifier appraise it?
  • How do isolation and attestation address distinct parts of a system’s threat model?
Back to all topics

Prototyping

System Manageability

I prototype manageability architecture for Arm systems, including baseboard management controllers (BMCs), the Security Protocol and Data Model (SPDM), and reliability, availability, and serviceability (RAS). This work connects management interfaces with platform security and fault handling.

  • BMC
  • SPDM
  • RAS architecture
  • OpenBMC

Questions & considerations

  • How should a BMC communicate with an SoC and other system components?
  • How should SPDM authentication and measurements be implemented?
  • How should a system report faults, contain their effects, and support recovery (RAS)?
Back to all topics

Founder · Linepine

Teaching & Advocacy

I founded Linepine, a platform that aspires to bring technology closer to children around the world. I have run multiple summer sessions and am now looking to evolve the curriculum to make AI more accessible. This work reflects my commitment to teaching and digital literacy, with the aim of reducing barriers created by unequal access to technology.

  • Linepine
  • Technology education
  • Digital literacy
  • AI literacy
Visit Linepine (opens in a new tab)

Questions & considerations

  • How can the curriculum serve children with different levels of prior knowledge and access to technology?
  • How can AI concepts be taught in ways that build understanding of both capabilities and limitations?
  • How can future sessions assess understanding and use that evidence to refine the curriculum?
Back to all topics

Research interest

AI Safety & Security

I’m exploring AI safety and the security of agentic systems from a systems architecture perspective. My interests include how agents are granted access to tools and data, how that authority is constrained, and how controls can be evaluated against unintended actions and adversarial inputs.

  • AI safety
  • AI security
  • Agentic systems

Questions & considerations

  • Which actions can an agent take, and what limits its authority over tools and data?
  • How can we evaluate whether controls hold under adversarial inputs and unexpected tool behavior?
  • What can an evaluation demonstrate about safety, and where does uncertainty remain?
Back to all topics

Research interest

Hardware-enforced AI Safety

I’m interested in how hardware-enforced isolation and access controls could constrain AI systems, and what attestation could tell us about those controls. This is an exploratory interest grounded in my platform security work. I participated in the Hardware Assurance Programme’s 2026 cohort and welcome discussions with researchers in AI safety and security.

  • Hardware isolation
  • Access control
  • Attestation

Questions & considerations

  • Which constraints can hardware enforce, and what assumptions do those guarantees depend on?
  • What can attestation establish about the enforcement environment and its configuration?
  • Which safety properties still require software controls or behavioral evaluation?
Back to all topics