{ "$schema": "https://clarkohlenbusch.com/schemas/resume-ast-v1.json", "nodeType": "ResumeAST", "id": "clark-ohlenbusch-resume-ast-root", "metadata": { "version": "1.0.0", "createdAt": "2026-09-09", "language": "en-US", "canonicalUrl": "https://clarkohlenbusch.com/resume.ast.json", "interactiveUrl": "https://clarkohlenbusch.com/resume", "pdfUrl": "https://clarkohlenbusch.com/clark-ohlenbusch-resume.pdf", "nodeCount": 43, "parserTarget": "LLM Reasoning Kernels, Static Semantic Parsers & Machine Scrapers", "privacyCompliance": "Sanitized for public web. No telephone or private address data." }, "root": { "nodeType": "DocumentNode", "id": "doc-0", "header": { "nodeType": "HeaderNode", "id": "header-0", "identity": { "nodeType": "IdentityNode", "id": "identity-0", "legalName": "Clark Ohlenbusch", "preferredName": "Clark", "headline": "Software Engineer & Researcher · Agentic AI Security, Voice Systems & Systems Architecture", "location": { "locality": "Boston", "region": "Massachusetts", "country": "United States", "timezone": "America/New_York (UTC-5)" }, "bio": "Software engineer, product lead, and security researcher operating across the entire stack—from bare-metal memory management (C/POSIX) and mathematical proof automation (Lean 4/Coq) to multi-agent zero-trust runtime governance (OPA/Rego) and low-latency acoustic conversational AI." }, "contacts": { "nodeType": "ContactGraphNode", "id": "contacts-0", "email": "clark.ohlenbusch@gmail.com", "website": "https://clarkohlenbusch.com", "github": "https://github.com/ClarkOhlenbusch", "linkedin": "https://www.linkedin.com/in/clark-ohlenbusch-bb8b60253/" } }, "sections": [ { "nodeType": "SectionNode", "id": "sec-thesis", "sectionKey": "operational-thesis", "title": "Operational Profile & Technical Thesis", "order": 1, "children": [ { "nodeType": "ThesisStatementNode", "id": "thesis-0", "statement": "Dedicated to designing high-integrity, deterministic, and verifiable systems in high-stakes environments. Operates at the intersection of agentic AI runtime security, full-duplex conversational voice systems, and bare-metal memory/proof verification." } ] }, { "nodeType": "SectionNode", "id": "sec-experience", "sectionKey": "professional-experience", "title": "Professional Trajectory", "order": 2, "children": [ { "nodeType": "WorkExperienceNode", "id": "exp-saviynt", "organization": "Saviynt Inc.", "organizationUrl": "https://saviynt.com", "role": "Product Manager & Forward Deployed Engineer", "division": "Agentic AI Security (Zuma Platform)", "dateRange": { "startDate": "2023-08", "endDate": null, "isCurrent": true, "display": "AUG 2023 – PRESENT" }, "location": "Boston, MA / Remote", "metrics": { "dealValueClosedWon": "$15M+", "enterpriseRfpsContributed": 22, "keyClientsEngaged": [ "Northern Trust", "FIS" ], "engineersTrained": 100 }, "achievements": [ { "nodeType": "AchievementNode", "id": "ach-saviynt-1", "category": "product-lifecycle", "summary": "Technical lead and product manager for Zuma through General Availability (GA).", "description": "Authored Builder Experience PRFAQs for agent SDK/CLI, promotion workflows, and secret scanning.", "techStack": [ "Zuma SDK", "CLI", "Secret Scanning", "Dev-to-Prod Promotion" ] }, { "nodeType": "AchievementNode", "id": "ach-saviynt-2", "category": "runtime-authorization", "summary": "Architected runtime authorization engine integrating OPA/Rego access control.", "description": "Implemented intent-deviation detection, behavioral monitoring, and led live technical POCs to enterprise clients including Northern Trust.", "techStack": [ "OPA", "Rego", "Behavioral Monitoring", "Intent Deviation" ] }, { "nodeType": "AchievementNode", "id": "ach-saviynt-3", "category": "governance-sod", "summary": "Defined enterprise Segregation of Duties (SOD) violation strategy for multi-agent chains.", "description": "Enforced design-time governance and runtime boundary controls across A2A (agent-to-agent) and H2A (human-to-agent) workflows.", "techStack": [ "Segregation of Duties", "A2A Security", "H2A Governance", "Zero-Trust" ] }, { "nodeType": "AchievementNode", "id": "ach-saviynt-4", "category": "commercial-execution", "summary": "Secured ~$15M in closed-won deal value across 22+ enterprise RFPs/RFIs.", "description": "Ran first live customer trial (FIS) structuring trial scope and conversion funnels.", "techStack": [ "Enterprise Architecture", "RFP/RFI Strategy", "Commercial POCs" ] }, { "nodeType": "AchievementNode", "id": "ach-saviynt-5", "category": "integrations-webhooks", "summary": "Led feature specifications for JWT-signed pre/post-tool webhooks and multi-cloud agent discovery.", "description": "Covered AWS Bedrock, Azure AI, GCP Vertex AI, Copilot Studio, and Salesforce Agentforce.", "techStack": [ "JWT Webhooks", "AWS Bedrock", "GCP Vertex AI", "Azure AI Foundry", "Copilot Studio", "Salesforce Agentforce" ] } ] }, { "nodeType": "WorkExperienceNode", "id": "exp-spoqen", "organization": "Spoqen", "role": "Co-Founder & Product Lead (Acquired via IP Asset Sale)", "division": "Voice AI Platform & Booking Agents", "dateRange": { "startDate": "2025-09", "endDate": "2026-03", "isCurrent": false, "display": "SEP 2025 – MAR 2026" }, "location": "Boston, MA / Remote", "achievements": [ { "nodeType": "AchievementNode", "id": "ach-spoqen-1", "category": "origin-and-leadership", "summary": "Spun startup out of 2nd-place North American finish at AstraZeneca Hackathon (80+ teams); led 5–6 person engineering team.", "description": "Led a distributed engineering team across weekly sprint execution, product discovery, and production architecture.", "techStack": [ "AstraZeneca Hackathon", "Team Leadership", "Sprint Planning", "Voice AI" ] }, { "nodeType": "AchievementNode", "id": "ach-spoqen-2", "category": "crm-booking-agent", "summary": "Deployed 24/7 autonomous voice reservation agent integrated with ResNexus CRM with 100% call-resolution metric.", "description": "Constrained CRM automation surface by 95% to achieve 65–75% autonomous completion with zero-latency human handoff.", "techStack": [ "VAPI", "ResNexus CRM", "Headless Browser Automation", "Call Resolution Metric" ] }, { "nodeType": "AchievementNode", "id": "ach-spoqen-3", "category": "multi-agent-compiler", "summary": "Architected multi-agent graph compiler (LangSmith + Gemini 2.5 Flash 1M context) parsing audio into visual DAGs.", "description": "Processed raw client call recordings to surface unmapped edge cases and slash voice workflow authoring from days to minutes.", "techStack": [ "LangSmith", "Gemini 2.5 Flash", "Multi-Agent Architecture", "DAG Canvas", "Transcript-to-Flow" ] }, { "nodeType": "AchievementNode", "id": "ach-spoqen-4", "category": "voice-infra-optimization", "summary": "Evaluated voice infra across VAPI vs Retell and Cartesia Sonic; optimized Claude 3.5 Haiku tool-calling prompts to ~15K tokens.", "description": "Migrated real-time event notifications to Pusher WebSockets with Sentry monitoring to reduce compute costs and ensure sub-second execution.", "techStack": [ "Claude 3.5 Haiku", "Cartesia Sonic", "Pusher WebSockets", "Sentry Telemetry", "VAPI" ] }, { "nodeType": "AchievementNode", "id": "ach-spoqen-5", "category": "bd-and-distribution", "summary": "Negotiated strategic partnership with Upside AI accessing 13,000 RV parks ($4,500/mo unit economics vs $50K payroll).", "description": "Anchored revenue share pricing against human front-desk replacement economics; advanced pilot discovery with Stack (~500 residential units).", "techStack": [ "Upside AI Partnership", "Unit Economics", "Enterprise Pilot", "Stack" ] }, { "nodeType": "AchievementNode", "id": "ach-spoqen-6", "category": "ip-asset-acquisition", "summary": "Concluded venture in March 2026 via successful proprietary IP and technology asset sale.", "description": "Transferred Spoqen's proprietary voice agent code, multi-agent compiler architecture, and booking pipeline to an enterprise acquirer.", "techStack": [ "IP Asset Sale", "Technology Acquisition", "Proprietary Codebase", "Venture Conclusion" ] } ], "outcome": "Concluded March 2026 via proprietary IP & technology asset acquisition" }, { "nodeType": "WorkExperienceNode", "id": "exp-meno", "organization": "Meno", "role": "Creator & Lead Architect", "division": "Socratic Voice AI Dialectic Learning", "dateRange": { "startDate": "2025-01", "endDate": "2026-01", "isCurrent": false, "display": "2025 – 2026 · BOSTON, MA / EDTECH VENTURE" }, "location": "Boston, MA (EdTech Venture)", "achievements": [ { "nodeType": "AchievementNode", "id": "ach-meno-1", "category": "founder-thesis", "summary": "Founded LearnMeno ('A more human way to learn') to reimagine learning and assessment through real-time Socratic voice AI.", "description": "Restored oral Socratic examination where students verbally articulate and deepen conceptual understanding against AI dialogue agents.", "techStack": [ "Socratic Pedagogy", "Dialectic Dialogue", "Voice-First Assessment", "Higher Education" ] }, { "nodeType": "AchievementNode", "id": "ach-meno-2", "category": "institutional-pilots", "summary": "Secured active university pilots at Boston College and UMass Boston with FERPA clearance for 250-student cohorts.", "description": "Cleared university IT security reviews, FERPA compliance constraints, and student AI-consent workflows (fall pilot committed at Colorado Mesa University).", "techStack": [ "Boston College Pilot", "UMass Boston Pilot", "FERPA Compliance", "IT Clearance" ] }, { "nodeType": "AchievementNode", "id": "ach-meno-3", "category": "audio-speech-pipeline", "summary": "Engineered real-time audio-to-audio dialectic assessment engine on Google Vertex AI (Whisper Turbo, Kokoro, Gemini Live).", "description": "Delivered sub-second turn-taking, natural interruption (barge-in) handling, and verbal comprehension scoring.", "techStack": [ "Google Vertex AI", "Whisper Large v3 Turbo", "Kokoro TTS", "Gemini Live API", "Barge-in Detection" ] }, { "nodeType": "AchievementNode", "id": "ach-meno-4", "category": "lms-integration", "summary": "Architected Canvas and Blackboard LMS roster integration with Google OAuth and tri-mode AI rubric grading.", "description": "Allowed students to take oral assessments via links with zero account creation friction; deployed professor dashboards with rubric evaluators.", "techStack": [ "Canvas LMS", "Blackboard LMS", "Google OAuth", "Rubric Engine", "Tri-Mode Grading" ] }, { "nodeType": "AchievementNode", "id": "ach-meno-5", "category": "gtm-and-capital", "summary": "Surveyed 30+ professors validating oral assessment as #1 pain point; pivoted GTM to department chairs; pitched SSC Venture Partners.", "description": "Pitched SSC Venture Partners ($10M fund); redirected product from practice tool to formal assessment and targeted Centers for Teaching & Learning.", "techStack": [ "Customer Discovery", "Professor Survey (n=30)", "GTM Pivot", "Venture Pitching" ] } ] }, { "nodeType": "WorkExperienceNode", "id": "exp-chai-lab", "organization": "Northeastern University (CHAI Lab)", "role": "Voice AI & HCI Researcher", "division": "Conversational Formality & Pacing", "dateRange": { "startDate": "2025-10", "endDate": null, "isCurrent": true, "display": "OCT 2025 – PRESENT" }, "location": "Boston, MA", "metrics": { "cohortParticipants": 140, "recruitmentPlatform": "Prolific" }, "achievements": [ { "nodeType": "AchievementNode", "id": "ach-chai-1", "category": "empirical-hci", "summary": "Deployed an empirical Voice AI trial (n=140 participants via Prolific).", "description": "Investigated how conversational formality, acoustic turn-taking pacing, and persona design influence trust, comprehension, and comfort in older adult cohorts.", "techStack": [ "Prolific", "Empirical HCI", "Acoustic Pacing", "Formality Metrics" ] }, { "nodeType": "AchievementNode", "id": "ach-chai-2", "category": "analytics-telemetry", "summary": "Architected real-time research dashboard tracking p-values, formality scores, and session audio playback.", "description": "Synthesized data for longitudinal analysis; preparing submission for ACM CHI / CUI.", "techStack": [ "Telemetry Dashboard", "Statistical Analysis", "ACM CHI / CUI" ] } ] }, { "nodeType": "WorkExperienceNode", "id": "exp-calico", "organization": "Calico Care", "role": "AI Development Consultant", "division": "Clinical Voice Outreach", "dateRange": { "startDate": "2024-10", "endDate": null, "isCurrent": true, "display": "OCT 2024 – PRESENT" }, "location": "Boston, MA", "achievements": [ { "nodeType": "AchievementNode", "id": "ach-calico-1", "category": "clinical-compliance", "summary": "Designed HIPAA-compliant voice AI architecture replacing legacy Alexa systems.", "description": "Utilized VAPI and encrypted AWS S3 audio storage for COPD/CHF patient outreach protocols.", "techStack": [ "HIPAA Compliance", "VAPI", "AWS S3", "Clinical Protocols" ] }, { "nodeType": "AchievementNode", "id": "ach-calico-2", "category": "clinical-evals", "summary": "Engineered automated clinical testing suite with 12+ scenarios.", "description": "Evaluated patient safety, regulatory compliance, and multilingual voice capabilities (Portuguese, German).", "techStack": [ "Clinical Evaluation", "Multilingual Voice", "Safety Guardrails" ] } ] }, { "nodeType": "WorkExperienceNode", "id": "exp-weymouth", "organization": "Town of Weymouth", "role": "Security Engineering Consultant", "division": "Enterprise Authentication", "dateRange": { "startDate": "2025-05", "endDate": "2025-08", "isCurrent": false, "display": "MAY 2025 – AUG 2025" }, "location": "Weymouth, MA", "achievements": [ { "nodeType": "AchievementNode", "id": "ach-weymouth-1", "category": "identity-security", "summary": "Designed and deployed hardware YubiKey MFA across Windows enterprise domain.", "description": "Integrated Active Directory and ADFS; verified end-to-end authentication workflows, authored runbooks, and trained IT staff.", "techStack": [ "YubiKey FIDO2/WebAuthn", "Active Directory", "ADFS", "Security Runbooks" ] } ] }, { "nodeType": "WorkExperienceNode", "id": "exp-umb-ta", "organization": "University of Massachusetts Boston", "role": "Teaching Assistant — CS240: Coding in C", "division": "Department of Computer Science", "dateRange": { "startDate": "2024-01", "endDate": "2026-02", "isCurrent": false, "display": "JAN 2024 – FEB 2026" }, "location": "Boston, MA", "metrics": { "studentsMentored": 100, "semestersTaught": 5 }, "achievements": [ { "nodeType": "AchievementNode", "id": "ach-umb-1", "category": "systems-mentorship", "summary": "Mentored 100+ students across 5 semesters in systems programming.", "description": "Taught manual heap allocation (malloc/free), pointer mechanics, segfault triage (GDB/Valgrind), and POSIX file I/O.", "techStack": [ "C", "Pointers", "Memory Management", "GDB", "Valgrind", "POSIX" ] }, { "nodeType": "AchievementNode", "id": "ach-umb-2", "category": "provenance-research", "summary": "Co-authoring research on Delegation Provenance Records (DPR).", "description": "Establishing cryptographic lineage tracking and runtime boundary verification for multi-agent chains.", "techStack": [ "Delegation Provenance Records", "Cryptography", "Multi-Agent Systems" ] }, { "nodeType": "AchievementNode", "id": "ach-umb-3", "category": "formal-verification", "summary": "Developed foundation model theorem prover pipelines synthesizing deterministic tactics.", "description": "Validated proofs using Lean 4 and Coq proof checker kernels with zero hallucination.", "techStack": [ "Lean 4", "Coq", "Theorem Provers", "Formal Methods" ] } ] } ] }, { "nodeType": "SectionNode", "id": "sec-education", "sectionKey": "education-and-honors", "title": "Education & Verified Honors", "order": 3, "children": [ { "nodeType": "InstitutionNode", "id": "edu-umb", "institution": "University of Massachusetts Boston", "institutionUrl": "https://www.umb.edu", "degree": "Bachelor of Arts in Computer Science", "gpa": { "score": "3.58", "scale": "4.00" }, "graduationDate": "May 2026", "honorsList": [ "Dean's List (5 Consecutive Semesters)", "John F. Kennedy (JFK) Award Nominee" ] }, { "nodeType": "AwardListNode", "id": "awards-ledger", "awards": [ { "nodeType": "AwardNode", "id": "award-jfk", "name": "John F. Kennedy (JFK) Award Nominee", "issuer": "University of Massachusetts Boston", "description": "University-wide award honoring students who embody civic engagement, intellectual achievement, and leadership." }, { "nodeType": "AwardNode", "id": "award-astrazeneca", "name": "AstraZeneca Global Health Hackathon Runner-Up (2nd Place Global)", "issuer": "AstraZeneca", "description": "Engineered vocal biomarker agent detecting subtle acoustic indicators of early-onset Alzheimer's." }, { "nodeType": "AwardNode", "id": "award-umb-ai", "name": "UMass Boston AI Symposium Hackathon — First Place & Crowd Favorite ($1,000)", "issuer": "UMass Boston", "description": "Built real-time respiratory acoustic classification system." }, { "nodeType": "AwardNode", "id": "award-hanopol", "name": "Elizabeth Hanopol Merit Scholarship", "issuer": "Department of Computer Science", "description": "Awarded for exceptional academic performance and technical promise." }, { "nodeType": "AwardNode", "id": "award-senate", "name": "Student Senate Merit Scholarship", "issuer": "Student Government Association", "description": "Recognizing instructional leadership and service to the student body." } ] } ] }, { "nodeType": "SectionNode", "id": "sec-skills", "sectionKey": "technical-taxonomy", "title": "Technical Skills Matrix & Taxonomy", "order": 4, "children": [ { "nodeType": "SkillDomainNode", "id": "skills-systems", "domain": "Systems & Low-Level Languages", "skills": [ "C (Low-Level Systems)", "POSIX APIs", "Memory Allocation (malloc/free)", "Valgrind & GDB Triage", "Rust", "TypeScript / JavaScript", "Python", "Shell & Bash", "SQL", "Lean 4", "Coq" ] }, { "nodeType": "SkillDomainNode", "id": "skills-security", "domain": "AI Security, Runtime Governance & Zero-Trust", "skills": [ "Open Policy Agent (OPA)", "Rego Policy Language", "Model Context Protocol (MCP)", "Zuma Platform Architecture", "Delegation Provenance Records (DPR)", "Multi-Agent Segregation of Duties (SOD)", "JWT Signed Pre/Post-Tool Webhooks", "Okta OAuth Token Exchange" ] }, { "nodeType": "SkillDomainNode", "id": "skills-voice", "domain": "Real-Time Multimodal Voice AI & HCI", "skills": [ "VAPI Telephony Integration", "WebSockets Audio Streaming", "Sub-Second Full-Duplex Audio", "Acoustic Turn-Taking & Pacing", "Natural Interruption / Barge-in", "HIPAA Compliance & Secure S3 Audio", "Visual Prompt Graph Canvas Architecture", "Prolific Empirical Trial Design" ] }, { "nodeType": "SkillDomainNode", "id": "skills-cloud", "domain": "Cloud, Infrastructure & Foundation Models", "skills": [ "AWS Bedrock & S3", "GCP Vertex AI", "Azure AI Foundry", "Microsoft Copilot Studio", "Salesforce Agentforce", "Docker Containerization", "Redis Caching", "Supabase & Postgres RLS", "Linux Administration" ] } ] } ] } }