Product
Build the product.
Fractional product and delivery leadership for founders who need a product team and for companies whose product lifecycle needs an owner, plus zero-to-one launches and turnarounds of stalled projects.
25+ years consulting and building products, including as President & GM at Axian, SVP Operations at Janrain, and VP Cloud Services at Aerospike.
Process
Build the process.
AI-assisted Spec-Driven Development and software development lifecycles (SDLCs) designed around your team’s own criteria, so AI-assisted delivery keeps the discipline production software needs and stays aimed at the outcomes your customers value.
Most product problems are also process problems.
We work on both: the product your customers use and the process your team uses to build it.
Services
For a founder who needs a product team, or the person who owns product development and the lifecycle at an established company. We take on a small number of clients at a time, own the roadmap and its delivery, develop your product managers, and align engineering, sales and marketing on the same outcomes, with Servo as how we work.
Decide where to play and how to win: market strategy, segmentation and buyer and user personas, positioning, pricing and packaging, and the launch plan that gets sales and marketing ready.
Take a new product from opportunity to launch: launch customers signed before the build, a scope cut to the first release, success measures agreed up front, and pricing tested against real demand.
For products that have stalled, drifted, or become unreliable: an end-to-end review, a decision on what to cut, and a sequenced plan to ship.
Bring engineering discipline to AI-assisted development: specs before code, approval gates, guardrails, evaluation, and cost caps designed in before deployment.
A product workflow set up around your criteria: team size, tooling, risk, and how much review each change needs. It comes with an audit trail of every run and metrics stakeholders can plan around.
Know where to direct your AI investment before you spend it: where to maximize your ROI, which approach fits the size of the problem, and where a human review step protects your customers.
That includes the model: offline, open-source models can cut costs for the right workloads.
Recent work
Process · Our method
A multi-agent framework that runs the product lifecycle inside the tools your team already uses, from defining a customer outcome to measuring it in production. GitHub, OpenSpec and Claude Code are supported today. Agents write and review the specs and the code, and a person decides at every gate. Building Signal and Servo itself, the median time from approved spec to merged code was 37 minutes during a two-week sprint.
Product · Multi-agent system
Conceived the product, set the roadmap, and wrote every spec.
A personal agent for job and contract searching, built as a multi-agent system. It scores each new posting against your real experience and shows who you know at the company, so you can land the role through a warm introduction instead of a blind application. You stay in charge: it never sends or applies for you.
Built and shipped with Servo.
Process · AI build
A customer lifecycle tool: self-service onboarding for the customer, and an internal account view for the vendor.
Customers point their own AI client at a hosted URL with their existing API key. There is nothing to deploy, and no customer data or production systems are accessed.
It returns a ready-to-submit cluster configuration from the customer’s workload inputs and refuses anything outside what customers may configure.
Product · AI-assisted support workflow
A support workflow demo built around the customer’s problem and the information the support engineer needs to solve it. AI does the reading and drafting; the support engineer decides what the customer sees.
Paste a ticket. It sets a severity, flags whether to escalate, and answers the customer’s technical questions from the product documentation, with citations, flagging anything the docs don’t fully cover. Halyard Cloud is a fictional managed database service.
Built with the Claude API, Qdrant hybrid search, W&B Weave evaluations and FastAPI.
Principles
Enterprise SDLC discipline doesn’t get thrown out for AI. It gets adapted, and adapting it well is the skill.
Track record
Work built in-house as SVP Operations at Janrain and VP Cloud Services at Aerospike.
Product · VP Cloud Services, Aerospike
$18M→$70M+
Aerospike ARR growth over nearly six years, with this service a contributor
Brought in to build the global professional services org, then turned customer demand into a new business:
99.99% uptime; SOC 2 and ISO 27001 certified.
Developer product · PLG · VP Cloud Services, Aerospike
300+
trial databases within three months of launch
Took over a stalled two-year DBaaS project and rebuilt it end to end:
Developer API · SVP Operations, Janrain
75% faster
average customer deployment
Mapped how delivery actually configured every customer, separated provisioning from configuration, and turned it into a public API.
Average deployment time dropped from 180 days to 45.
SaaS platform · SVP Operations, Janrain
18→2
platform service events a year, cut in 12 months
A heatmap of critical components, the least stable microservices rebuilt, new features paused for six months, and every customer briefed on the root cause.
99.99% uptime.
Writing
Articles and a LinkedIn series on adapting enterprise SDLC discipline for AI-assisted development, written as systems are being built.
Real decisions, real mistakes, and the questions they raise. Everything here comes from Tom’s own projects or public information. He doesn’t name clients or share anything from their work without their permission.
“Once building is that fast, ‘how fast can we ship it’ becomes the easy question to answer. The harder question that’s left is the one it was always supposed to be: did the work matter to the customer?”
Read the articleMore from the series
On leadership
What people say
“I watched Tom turn around a demoralized team, I watched him turn burn out to purpose, I watched him replace failure with efficiency, and I watched him turn frustrated clients into committed partners.”
“Some of the projects we had were small, some quite large and critical to our business. Tom is someone I could always count on, someone who drove results through his teams and responded to our every need.”
“Characteristics that stand out to me are strategic thinking/execution and organizational (and personal) development… He launched a new practice, was not afraid to address a faltering one…”
About
Tom has spent more than 25 years consulting and building products and the processes behind them, starting as a Senior Product Manager at Equilibrium in 2000, then running product teams at Axian, heading delivery and platform operations at Janrain, and leading the cloud product lines at Aerospike.
He has launched five products from zero, rebuilt a SaaS platform to 99.99% uptime, and turned around a stalled developer product by rebuilding its team, architecture, and go-to-market.
He founded Switchfoot Advisory Group in May 2024. Today he works on product strategy and builds production AI systems with Spec-Driven Development, helps companies speed up their development teams with AI-assisted tooling, and takes on fractional product and delivery leadership engagements.
Need the product built?
Launch it, rescue it, or lead it.
Need the process built?
Make AI-assisted delivery production-grade.