Product Manager bridging business and AI.
I'm a Technical Product Manager with 5+ years bridging business and technology teams to ship AI-first, data-driven platforms. My work spans LLM integration, RAG pipelines, agentic workflows, and API-driven integrations — always with a prototype-and-iterate mindset. I'm currently filing a patent with my CTO for a 5-layer multi-agent AI platform, and I hold a Masters in Engineering Management from Cornell (3.88 GPA, full merit scholarship) and am pursuing a Doctor of Business Administration at Westcliff University.
Where I'm strongest as an AI product leader — self-assessed and backed by shipped outcomes.
Real products. Every decision documented.
From multi-agent AI platforms at Likewize to product strategy and 0→1 healthcare. Click any to explore the trade-offs and outcomes.
AI Launch Scorecard
A framework I use before shipping AI features. Try it.
A repeatable operating system for AI products.
The same five-step process I run from 0→1 — designed for regulated, high-stakes domains where being wrong is expensive. Every step produces a concrete artifact.
Frame the problem & the failure mode
Before a single prompt, I define what success looks like and — more importantly — what a wrong answer costs. In insurance and healthcare, a hallucination is a compliance event, not a UX bug. The risk surface defines the spec.
Prototype to learn, not to impress
I build a working slice fast with Lovable, v0, and Claude tooling, then put it in front of real users and execs within days. A live demo surfaces the real requirements that no discovery doc ever will.
Design the eval before the model
I write the evaluation rubric — accuracy, safety, latency, explainability, recovery — before choosing an approach. Weighted scorecards turn 'feels good' into a defensible ship/no-ship decision.
Keep a human in the loop
Agents do the heavy lifting; people make the call with a visible audit trail. This is what flips adoption from 'interesting' to 'can't work without it' in regulated environments.
Instrument, monitor, iterate
Every AI feature ships with telemetry: confidence distributions, deflection, escalation, drift. The roadmap is driven by what production tells me, not what the backlog assumes.
MCP × n8n. Production workflows.
Lead Qualification Pipeline
Inbound HubSpot form submissions trigger an n8n workflow: MCP enriches company data via LinkedIn + Clearbit, Claude scores lead quality (1-100) with reasoning, auto-routes hot leads to a Slack channel, and logs to Airtable.
Contract Intelligence Pipeline
Legal uploads PDFs to Google Drive. n8n detects new files, MCP extracts text via Document AI, Claude identifies key clauses (termination, liability, auto-renewal), compares against standard terms, and pushes a risk-scored summary to Notion.
Customer Feedback Loop
Pulls NPS from Delighted, tickets from Zendesk, reviews from App Store Connect via MCP. Claude runs sentiment analysis, clusters themes via embeddings, and autopublishes a weekly product insight report to Confluence.
Competitive Intel Monitor
Monitors 8 competitor sites + Product Hunt + Crunchbase via MCP scraping. A diff engine detects pricing changes and new features. Claude writes a daily brief against our roadmap positioning, delivered 7am to Slack.
Start with the failure mode
Most PMs ask 'what should the AI do?' I ask 'what happens when it's wrong?' In regulated domains like insurance and healthcare, wrong answers are compliance violations, not bad UX. Designing guardrails IS the product work.
Trust is the product
In regulated environments, adoption flips the moment a human stays in control. My Traceability agent only took off once I added the sign-off loop. Agents do the heavy lifting; people make the call with a visible audit trail.
Prototype to persuade
A working demo beats any deck. I use Lovable and Claude tooling to turn abstract LLM orchestration into interactive prototypes — which is how I've secured executive sponsorship for AI-first initiatives.
Where I've worked.
Likewize Device Protection
- ▸Filing a patent with the CTO for a 5-layer, multi-agent AI platform that transforms unstructured legal documents into production-ready insurance portals — covering T&C-to-portal generation, compliance-embedded AI, and dynamic knowledge graph configuration.
- ▸Built a closed-loop AI agent ecosystem (Discover → Build → Review → Validate) that cut requirements cycle time ~50%. The Traceability agent alone eliminated 8+ hours of manual BRD work and surfaced coverage gaps across 50+ requirement line items before every release.
- ▸Defined functional features for a fintech B2B SaaS platform targeting $500K+ ARR, serving 7M+ global users across banking clients (Santander, Barclays, Stripe, Access PaySuite) in EMEA.
- ▸Owned roadmap prioritization via impact-risk-effort analysis across 40+ features with 2 teams (12 engineers & QAs); selected by SVP to represent Product in executive funding reviews across 3 UK offices.
Sam Analytic Solutions (eBoxchain)
- ▸As the first product hire, raised $250K in funding by presenting eBoxchain's solutions to VCs and articulating unique value propositions.
- ▸Defined requirements for a HIPAA-compliant platform with 8+ third-party API integrations (pharmacy management systems), evaluating risks, scalability, and regulatory readiness from 0 to 1.
- ▸Specified CRM integration requirements and led a team of 6 — delivered 19% improvement in operational reporting and converted 4 new clients within 3 months.
MŌD Advisors
- ▸Designed data analysis workflows using vector databases (Pinecone), embedding models (Azure OpenAI, Anthropic Claude), and RAG pipelines; interviewed 15+ SMEs to prototype AI-driven solutions for retail and financial services clients.
- ▸Conducted competitive & market analysis across 10+ market players; synthesized findings into C-suite presentations informing $12M in annual cost savings and 13% ROI improvement on CapEx.
Johnson & Johnson
- ▸Launched the Product Lifecycle Management (PLM) System campaign for collaboration between strategic supply chain partners — increasing product impact by $2M in a year.
- ▸Conducted customer interviews across aerospace, automotive & defense industries; owned the project roadmap and consolidated 3 PLM systems into one.
Dell Technologies
- ▸Managed an enterprise IT infrastructure portfolio serving financial services clients (Vanguard, USAA) across North America; improved digital transformation objectives by 35%.
- ▸Resolved 145+ mission-critical escalations, improving SLA metrics by 4% each, NPS by 7%, and elevating pipeline value by ~$900K per account.
Westcliff University
Cornell University
Mody University
Let's build something
that actually ships.
Open to AI Product Manager roles at ambitious companies.
nj222@cornell.edu · +1 (607) 595-8222 · Irving, TX