Technical founder and full-stack AI builder with seven years inside the risk and infrastructure stack of top-tier financial institutions (Goldman Sachs, BlackRock, DTCC) and a Quantitative Finance degree. Currently running TipsData on a 15-agent virtual operations stack built on the paperclip.ing runtime — agents personifying The Office cast covering CEO/CFO/CMO, sales (inbound and outbound), data and security engineering, QA, financial modelling, design, technical writing, and social — driving a real business over a 1.2B-record cloud-native data platform. Operates idea-to-production in two weeks, speaks the language of risk and PnL, and ships directly to customers without intermediaries.
Core Competencies
Multi-Agent Virtual Operations
Production LLM Systems
Agentic Workflows & Tool Use
Skill / Plugin Architecture (paperclip.ing)
Financial Systems & Risk Tech
Kubernetes & Cloud-Native Infra
Forward-Deployed / Customer-Facing
Fast Prototyping (Idea → Prod)
Work Experience
TipsData / Database Emailer — Jan 2026 – Present
Founder / CEO / CTO — Multi-Agent Operations on a 1.2B-Record Data Platform
Run the business on a 15-agent virtual operations stack built on the paperclip.ing skill/plugin runtime — agents personifying The Office cast covering CEO, CFO, CMO, outbound sales, relationship sales, customer support, data engineering, security engineering, QA, financial modelling, design, technical writing, and social analytics
Designed the skill and plugin surface each agent uses — repo and dataset exports, ClickHouse / Postgres queries, email-delivery actions, CRM hooks — so agents act on real systems instead of just chatting
Operate paperclip self-hosted on a private VPS with isolated workspaces and instances per agent driven over its HTTP API; iteration cadence is hours, not sprints
Replatformed the legacy on-prem LAMP stack to a cloud-native architecture on Kubernetes (Vultr / AWS), ClickHouse, Neon PostgreSQL, React, and Node.js
Built the query and export engine handling 1.2 billion contact records with sub-second response on filtered columnar reads
Shipped email-delivery infrastructure on Microsoft 365 and AWS SES; integrated HyperDX, PostHog, GTM, and GA for end-to-end identity-resolution analytics
Own full CI/CD in GitHub Actions — image build, scan, and rolling deploy across the platform and the agent stack
Built and shipped an AI documentation platform with a Python engine running on Anthropic Claude, OpenAI, and Ollama behind a single multi-provider abstraction layer — model-agnostic routing, fallback, and per-customer policy
Designed a multi-tenant SaaS on DynamoDB single-table (6 GSIs for tenant isolation), RBAC, and real-time collaborative editing via Liveblocks + Yjs; built async LLM pipelines on Redis/BullMQ + AWS EventBridge with per-tenant rate isolation
Integrated 4 Git providers (GitHub, GitLab, Bitbucket, Azure DevOps) with OAuth and SSH deploy keys; deployed on Vultr VKE after an earlier AWS Lambda + DynamoDB iteration under Terraform
Sunset the product in April 2026 after concluding that high-quality documentation generation requires human context the agent could not procedurally infer — kept the open-source core (parses 7 languages, generates 10+ Mermaid diagrams, ships Docusaurus sites); the lesson directly informed the human-in-the-loop, role-specialized agent design now running at TipsData
Work Experience
Tooths — Aug 2023 – Present
CoFounder / CTO — Dental Health Marketplace
Architected a three-sided marketplace (patients, hygienists, dentists) on AWS EKS microservices backing React Native and React web clients
Built Helm charts with isolated dev/staging/prod environments and CI/CD pipelines that gate on code scanning
Stood up DataDog observability with Slack alerting and Vanta for automated compliance evidence collection
BlackRock — Jan 2022 – Aug 2023
Associate — Infrastructure Technology Audit · New York, NY
Led infrastructure technology audits across the Aladdin / portfolio platform stack — Change Management, Access Controls, Version Controls, Job Monitoring
Coverage spanned databases, observability, and message queues; built fluency in how a top asset manager actually runs investment systems in production
Led audits over distributed databases, data loss prevention, VMware virtualization, access controls, change management, and IT asset management at one of the most systemically important financial market utilities
Provided technology assurance over Market, Credit, Liquidity, and Operational Risk tech under Dodd-Frank and CCAR — direct exposure to the risk engines hedge funds and banks use to size positions
Automated workpaper generation against Goldman's SecDB / Slang API, replacing manual evidence collection for Job Monitoring, Change Management, Version Controls, and Data Integrity
Tips Marketing Services — Dec 2018 – Jan 2022
Consultant — Infrastructure
Maintained co-location environment and Linux production servers while designing the cloud re-platforming strategy that later became Database Emailer / TipsData
Selected Projects
PRODUCTION
TipsData — 15-Agent Virtual Operations
A real business run by a fifteen-agent stack on paperclip.ing, each agent personifying a character from The Office and owning a distinct function. Each agent has its own isolated workspace and a tool surface: ClickHouse/Postgres queries, dataset exports, email-delivery actions over SES and Microsoft 365, and CRM hooks. Self-hosted on a private VPS, acting on a 1.2B-record cloud-native data platform.
Cloud-native data platform serving query, export, and delivery over 1.2B contact records — replatformed from legacy LAMP to Kubernetes + ClickHouse + Postgres with end-to-end observability.
Multi-provider AI documentation platform (Claude / OpenAI / Ollama behind one abstraction) with async LLM pipelines, multi-tenant isolation, and direct Git-provider integration. Sunset April 2026; open-source Python core remains public.
Anthropic Claude API · OpenAI API · Ollama · Python · Node.js · DynamoDB · Redis/BullMQ · AWS EventBridge · Liveblocks · Yjs · Terraform