Darren Card · VancouverFull-time or fractional
Everyone can ship an agent now. Far fewer can read across the systems a team already runs and say what the work returned.
Twenty years building and leading product and technology teams, most of it as the first product leader in the room. Available for a director-to-chief seat, or fractionally, on eight fixed-purpose missions.
Darren Card
Open to senior product leadership at B2B SaaS and AI companies where product is the business. Vancouver, on Pacific. Ran product for an Eastern-time company from here for three years.
Product leader, builder, operator · Ships production AI · Founding team at AI-native startup · B2B SaaS expert · 4x category creation
- Board and investor rooms
The product story behind a $5.5M Series A, and a quarterly advisory board with Microsoft, Box and Red Hat.
- Stage range
Employee #1 at pre-seed, first product and technology leader at Series A, VP through $10M to $50M ARR.
- Brought in from outside
Recruited by the CEO as the first product leader in the room, five times.
- Vertical breadth
Eight verticals shipped in, including commerce infrastructure, martech, learning and HR tech.
- Founder
Co-founded Popgun Media and ran it, 1999 to 2006: the business, the development and the delivery.
- AI in production
LLM, RAG and MCP shipped to paying customers, with the unit economics owned end to end.
The long version
Product and technology under one accountability. Most recently founding CPTO at Lexful, taking an AI-native platform from zero to paying customers in six months: the permission model governing what agents could read and act on, an MCP server so outside agents could call in under the same rules, and the evaluation framework that gated every release, with human review on the paths where a wrong answer costs a customer money. Every SOC 2 Type II and ISO 27001 control in place at launch with the audit underway, and the AI economics owned end to end: token budgets, inference tiers, and credit-based pricing that held margin as usage grew.
Before that, first product and technology leader at Cognota, where the SMB-to-enterprise pivot won EY, Synchrony Financial and Delta Air Lines and helped land a $5.5M Series A, and engineering moved from low to high on Google's DORA benchmarks in six months. Twenty years and eight B2B SaaS verticals behind it, from Vancouver, BC.
Full-time, or fractional. The conversation comes first.
Fill the seat outright, or bring me in fractionally. The fractional side runs as eight fixed-purpose missions, the diagnostic leads, and every one of them is scoped on a call before anything is committed.
A company filling a product, product operations, or product-and-technology seat, from director to chief level.
Product strategy and operations, owned end to end: a leader who builds and operates.
The seat filled by someone who has gone employee number one through post-Series B.
Twenty years in B2B SaaS across eight verticals, chief product and technology roles, zero-to-one launches through post-Series B scale. The recent record, scored, is further down this page.
An exec team that needs senior product and technology judgment in the room every week.
A standing seat on the AI agenda: portfolio, pricing and packaging, org design, measurement, hiring, board communication.
Senior judgment in the room every week, and it is reversible.
Runs as a standing seat, or as one of the eight fixed-purpose missions below.
Find out where you stand
The diagnostic · start hereA leader with several AI efforts running and no shared read on them.
Your AI work inventoried and ranked, the cost model, the measurement gaps, and a stop list.
You can name your bets, their costs, and the ones you stopped.
Put a number on good enough
A team where AI already produces the work, and no one can say whether the output is good enough to put in front of a customer.
Your quality bar pulled out of your own approval history into a scored eval, one candidate workflow measured against it, and the loop that closes when the score moves.
An eval your reviewers run on real work, and a decision they changed because of it.
Name the category you are in
A product buyers cannot place, in a market where AI has renamed everything.
The category named, the buyer it addresses, the language sales uses, and the proof the claim needs.
A category you can defend in a room, and a sales team using its language.
Take an AI product to market
A product leader with AI in the product and unknowns underneath it.
Customer conversations first, then margin per feature, packaging and pricing, and the route to value.
A priced roadmap, a modelled margin, and a story sales can defend.
Install the Agentic Product Operating Model
A software product org where agents produce much of the work and the weekly cycle has not moved.
Your teams classify their own work: what moves to agents on a schedule, what keeps a named owner.
One team running the new model, and the next phase chosen.
Build the team the work now needs
A leader hiring into a product org while agents change what each role is for.
Roles, levels and ratios re-cut against what agents now produce, the bar rewritten for judgment, and an interview loop that tests for it.
An org chart you can defend, a hiring plan naming the roles you stopped opening, and a loop that has run.
Bring AI into how the company operates
An established firm that sells expertise: engineering, environmental, professional services, field work.
One workflow first, not the whole firm. The expertise the work depends on becomes an agent that drafts, with sign-off kept human.
An agent drafting real work in production, and the next workflow chosen.
Go from idea to launch
A founder or founding team taking an AI-native product from nothing to market.
The whole zero-to-one arc: the raise, the founding team, the architecture, the build, the launch.
Live, with paying customers, and the operating model stays with the team.
With the harness: a named owner on every call, reporting your exec team can read without me in the room, and evals underneath that keep the numbers straight. And with a verdict, scored in three words, no others.
Whatever we automate, one person with a name still owns every call.
Clients push back on this line first and rely on it longest. Agents read, draft and check; the approval never happens without a person. The same rule applies to me: your teams classify their own work.
- Business fit
- Design quality
- Technical fit
- Final approval
Judgment. Never delegated, never automated.
- Read the context
- Draft the options
- Check the rules
- Propose the next move
Volume. Fast, tireless, and never the owner.
Everything above the rule has a name attached to it.
One client at a time, and a written definition of done at the start.
- Validation
You have already decided and want it confirmed. I will tell you what I think, which is occasionally unwelcome.
- A document
You want a strategy paper rather than working practice. Everything here ends with something running.
- No appetite to change
AI moves the work around. If the team does not want that, the engagement fails whatever the technology does.
- The cheapest option
I am not it, and pretending otherwise would waste your quarter and mine.
They had ninety people and one estimating desk everybody waited on.
An environmental and resource consultancy, mid-modernisation. This is the shape most of them take, and the reason the first move is never the whole company.
- 01Where it started
One senior specialist held the estimating judgment the whole firm queued behind. Every quote waited on the same desk, and so did the work behind it.
- 02What we did
That judgment became an agent that drafts. The sign-off stayed with a person, and heritage and First Nations data was excluded from extraction before a line was written.
- 03Where it stands
Fixed-fee discovery, with the success condition agreed before any build. Nothing is delivered yet, so the record says watch and shows no number.
The recent record, scored the same way I score everybody else.
Each entry had a definition of landed agreed before it started. The twenty years behind these are summarised below.
- 2021 to 2024LandedVP, product and technologyCognota, LearnOps platform for corporate L&D, Series AThe call
Recruited by the CEO as the first product and technology leader, running a 25-person product and engineering org, and led the SMB-to-enterprise pivot that defined the category.
What happened nextFour-figure ACVs became six-figure multi-year enterprise deals with EY, Synchrony Financial and Delta Air Lines, and the $5.5M Series A closed in December 2023.
- 2025 to 2026LandedEmployee #1, CPTO, founding teamLexful.ai, AI-native knowledge platform for MSP and IT teams, pre-seedThe call
Helped lead the pre-seed raise and hired the founding product and engineering team before the build started, then took the product from nothing to launch in about six months, with agents running in production under scoped permissions.
What happened nextThe number was paying customers, and it moved from zero to one on launch day, 4 February 2026, with every SOC 2 Type II and ISO 27001 control already in place and the observation window running.
- 2026 to nowWatchBuilder, independent product and technology R&D on AI-native ProdOpsDacard.ai, applied research, no outside capitalThe call
Built the scoring system and Throughline, the console that runs it, so the model had to survive contact with a real company.
What happened nextThroughline runs against one company. A sample of one, so it says watch.
- 2026 to nowWatchAI modernisation, discovery and designEnvironmental and resource consultancy, around ninety staff, client not namedThe call
Brought in mid-modernisation to find the first move and design it, with the scope and the definition of done agreed before any build.
What happened nextIn discovery, nothing delivered yet, so it says watch and shows no number.
Twenty years sit behind those four entries.
The background
- The span
Twenty years in B2B SaaS across eight verticals, deepest in MSP and IT services, martech, learning and HR tech, commerce.
- The seats
Chief product and technology roles, VP and director of product, and an operating partner seat with a venture firm.
- The names
Cognota, Elastic Path, Unbounce, Allocadia, QuickMobile and Vision Critical, through zero-to-one launches and an SMB-to-enterprise pivot.
- The categories
Four, named and taken to market: composable commerce at Elastic Path, marketing performance management at Allocadia, LearnOps at Cognota, Knowledge Intelligence at Lexful.ai.
- The caveat
Almost none of it had a definition of landed agreed up front, so it is summarised here rather than scored. References for any of it, on request.
Tell me what you are trying to find out.
One real question, worked live. You leave with a straight read on it, and a clear answer on whether I am the person to take it further.