Selected work

Work that shipped.

Renewals tracking for an early-stage cybersecurity firm selling to enterprises comes first, because it's the system a Seed or Series A team needs once the first enterprise contracts are signed. The rest is client work and work done at five employers, with a sailing game built solo at the end.

Work done at Boston Dynamics, iAdvize, Sheprd, Uber, and a sales platform for real estate belongs to those companies. A client appears by name only when the client has agreed to it. SailMOB is Ventus Optera's own product.

Projects

Design and build · Client work · 2026

Renewals tracking for 34 enterprise accounts, at an early-stage cybersecurity firm selling to enterprises

A small team at an early-stage cybersecurity firm selling to enterprises tracks renewals for 34 enterprise accounts in this system. Before it, the signed contracts sat in inboxes, shared drives, and spreadsheets, so the terms agreed with a client and the way that client was billed depended on who you asked. Every account now shows the contract in force, the signed documents behind it, and how the client is billed. The firm manages its accounts here, and at rollout the system had 31 client accounts and hundreds of contacts.

Change how a client is billed and last quarter's arrangement stays on record, in a change history that can't be quietly rewritten, even by an administrator. The firm's CEO priced a CRM license plus an implementation partner, then chose to own the system instead.

RenewalsClient accountsContract managementChange historyAccess control

Related offers: Internal tools and CRM and pipeline.

Illustration: contracts scattered across an inbox, a shared drive, a spreadsheet, and someone's memory, gathered into one client account page showing the contract in force, signed documents, billing terms, and history
Illustration: from five places to one page per client
Founding Director of Operations and Product · 2026

Operations and product at a sales platform for real estate

The product went from MVP to live beta in 2026, with operations and product run by its Founding Director from inside the founding team.

The parts of that work shown here are the research pipeline, the uptime page, and the vendor status page. The company stays unnamed on this site, and so does its product data.

OperationsProductResearch pipelineUptime pageVendor status

From MVP to live beta in 2026.

Product and build · Shipped at Boston Dynamics

Boston Dynamics Support Center

The rebuild of support.bostondynamics.com ran from a content audit through launch, with a new design system and one site connected to Salesforce, Jira, and Paligo.

Product, Services, and Training all publish into it under ITAR and US and EU compliance, and export rules travel with the content so that each reader sees what they're cleared to see. The Customer Learning and Development team operates the Docebo LMS and the Paligo CCMS, both connected to the site.

SalesforceConfluenceJiraFigmaDoceboPaligoGoogle AnalyticsBitlyEnterprise integration

Related offers: internal tools and CRM and pipeline.

The Boston Dynamics Support Center homepage: a search bar under the title, a registration banner, and product cards for Spot, Orbit, and Stretch
The shipped support center homepage
Client work · Agency finance

DealMap: expense allocation for agencies, built on Xero

The agency allocates each expense inside the Xero books it already keeps, so there's no second tool to maintain.

XeroExpense allocationAgency finance

Related offer: Internal tools.

One set of books, in Xero.

Systems review · In progress

Cutting SaaS sprawl for a Boston nonprofit sailing school

The school's operations are spread across 16 platforms, including Salesforce, scheduling, payments, assets, and maintenance tracking. The engagement is still running, and results go here once there are some.

Systems reviewSoftware selectionNonprofit

Related offer: the Ops Audit.

16 platforms under review.

Live chat sentiment analysis · At iAdvize

Galaxy S20 launch, for Samsung

The work at iAdvize was advising on and helping build a sentiment analysis program for Samsung's live chat during the Galaxy S20 launch. It flagged at-risk customers in real time and alerted senior executives.

OtterBox and LifeProof were advised customers at iAdvize as well.

HubSpotiAdvizeSentiment analysisLive agents

Read the success story. Related service: internal tools.

Samsung iAdvize chat box overlay on a phone screen
The iAdvize chat box on a phone screen
Trust and safety, growth, and onboarding · 2018

Sheprd

Sheprd is a rideshare service for kids, inside a regulated K-12 environment. The role owned trust and safety outright, from driver vetting and background checks to incident response and the compliance that let the service run. It also meant driving kids on a Massachusetts school bus license and working directly with parents, schools, and afterschool programs.

On the commercial side it owned the funnel, marketing automation, and cold calling. A rebuilt onboarding tripled new bookings and doubled rebookings.

Trust and safetyDriver vettingK-12 complianceOnboardingPartnerships

Related service: data cleanup and integrations.

Investigations and account security · 2014-2017

Uber

Investigations at Uber started as part of one role and grew into a 130-person program across two sites. Account-takeover response for North America was designed, built, and owned inside it. Better analytics and a redesigned escalation process cut time-to-solve on critical security escalations by half.

A support agent built R2D2, an internal Slackbot, in Python on a Slack integration, before LLMs. It answered about 85% of support agents' questions from the knowledge base, which the role owned along with the policy and process in it, and the role sponsored R2D2 itself. The logic for automated cancellation refunds was written there too.

PythonMySQLSlackConfluenceZendesk

Related service: internal tools.

Portrait of R2D2 with a lace ruff, representing the chatbot
R2D2, the knowledge base Slackbot
Browser sailing game · Own product · Since 2024

SailMOB

SailMOB is a browser sailing game with real wind physics, raced head to head against an AI fleet.

It was built solo with a sim-driven, rules-based approach, which means the simulation and the written rules come first and AI agents build against them. A technical founder curious about building fast with agents can read the method in the case study.

React 19CanvasSupabaseStripeAI bots
SailMOB 53 seconds before the start: the player's boat on a close reach below the start line, bots 2, 3, and 5 maneuvering nearby, live instruments, wind shadows, the no-go zone, and the start marker
SailMOB in the start sequence, racing the fleet
Next step

Which of these looks most like your problem?

Bring that one to the call, along with what's different about your company.