Rao Mikkilineni

Dr. Rao Mikkilineni


Research Focus & Metrics

Representative Publications

View All Publications
11 September 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Rao MikkilineniRao Mikkilineni
Rao Mikkilineni
WKW. Patrick Kelly
W. Patrick Kelly
Agentic AI moves enterprise governance from output review to authorized action. Contemporary state-of-the-art systems now provide pre-action approvals, policy gateways, durable workflows, identity controls, checkpointing, memory, and tracing; the unresolved systems question is therefore not governed versus ungoverned AI, but whether institutional authority itself survives scaling, failure, retry, migration, and executor substitution. This article presents a Mindful Machine architecture in which a ratified Digital Genome encodes purpose, delegated authority, policies, invariants, non-functional requirements, and recovery duties; the Autopoietic and Metacognitive Orchestration System (AMOS) mediates consequential action; an Authorization Warrant carries state-bearing authority; and one authoritative event history supports service-delivery, assurance, and governance views. A synthetic prior-authorization implementation provides the empirical anchor. A read-only query of current graph completion records yielded 25 cases: 10 accepted (40%), 3 denied (12%), and 12 escalated (48%), exercising all three designed authority paths. Escalation is interpreted as explicit transfer of decision custody to accountable human review rather than autonomous commitment. The results demonstrate governance-path execution and completion semantics, not clinical efficacy. Future work extends decision closure to consequence closure across human reviewers, providers, patients, appeals, and other external actors.

3 November 2026 in Information
Verified
Journal Article
Open access
PDF file
Rao MikkilineniRao Mikkilineni
Rao Mikkilineni
Increasingly capable AI systems can generate proposals, invoke tools, and create institutional commitments. Their cognitive performance does not establish their authority to act. This conceptual systems article develops governed systemic intelligence as an assurance framework for human–machine systems, rather than redefining intelligence to include governance. It distinguishes capability, epistemic justification, operational authorization, and normative legitimacy. A structured synthesis of intelligence, augmentation, cybernetics, computer security, provenance, and AI oversight motivates eight architectural requirements: ratified purpose, capability–authority separation, bounded authorization, authoritative memory, consequence closure, ratified learning, governed continuity, and external verifiability. A unified traceability matrix connects each requirement to an evaluation criterion, evidence artifact, and omission counterexample. Predicates over an explicit system model separate governance conformance from measured augmentation and from the stronger claim of domain-superhuman performance. A worked hypothetical prior-authorization episode illustrates all eight requirements, while a numerical coherence-debt example makes unresolved discrepancies and missing outcomes visible. Mindful Machines and a conventional policy-and-workflow stack provide alternative implementation mappings. Previously reported implementations support component feasibility and governance-path execution but do not establish complete conformance, scalable oversight, or superhuman capability. The contribution is an assurance-case pattern that composes authority, memory, continuity, and correction obligations into testable claims, with explicit limits under adversarial behavior, institutional capture, and incomplete consequence observation.

19 October 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Rao MikkilineniRao Mikkilineni
Rao Mikkilineni
WKW. Patrick Kelly
W. Patrick Kelly
Enterprise agentic AI requires an architecture that preserves institutional authority while models, tools, and execution resources change. This article develops the Mindful Machine as a governance-native process architecture in which a ratified Digital Genome specifies purpose, delegated authority, constraints, expected consequences, and recovery duties. The Autopoietic and Metacognitive Orchestration System executes declared governance processes; consequential actions require bounded Authorization Warrants; and an episode of record links proposals, decisions, effects, observations, obligations, and accountable closure. Eight properties make the architecture testable: purpose governance, authority separation, continuity preservation, consequence grounding, human-accountable closure, self-regulation, causal memory, and substrate independence. A conditional argument establishes commitment admission under explicit enforcement assumptions, while a conformance method separates architectural requirements from implementation evidence. Published video-streaming and credit-default implementations establish a broader engineering lineage; governed account recovery illustrates consequence recording and closure; and 25 synthetic prior-authorization records contain 10 accepted, 3 denied, and 12 escalated dispositions. These implementations provide existence evidence for reusable mechanisms rather than a universal assurance claim. The contribution is a coherent specification of authority custody and commitment-aware continuity across heterogeneous workflows, together with falsifiable requirements for evaluating its realizations.

24 September 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Rao MikkilineniRao Mikkilineni
Rao Mikkilineni
Superintelligence is usually defined as a machine intellect that exceeds human cognition across most domains. That framing conflates three distinct properties: computational capability, warranted knowledge, and legitimate authority. This article develops a systems definition suited to large language models, tool-using agents, databases, theorem provers, and other computational knowledge artifacts. Using conceptual synthesis and constructive requirement derivation, we distinguish a computational artifact, an operational self that preserves commitments and consequence history, and a normative principal that holds purpose and accountability. Human-created programs, proofs, databases, and learned models can embody accumulated knowledge and recorded consequences without phenomenal experience; what converts stored information into operationally discernible knowledge is provenance, validation, consequence closure, and revisability. We define superintelligence as a property of a constitutionally governed human-machine system that applies knowledge beyond unaided human scale, speed, or reliability within a declared scope while preserving human-ratified purpose, independently bounded authority for consequential action, authoritative memory, and ratified learning from outcomes. Eight architectural properties and falsifiable evaluation criteria are derived. Published Mindful Machine implementations are used as constructive feasibility evidence, while current regulatory and runtime-governance frameworks delimit what remains to be demonstrated.