Operations and AI Systems Consultant

Lead handoffs, ops workflows,
and AI where it pays.
Built and shipped
in weeks, not quarters.

I design and build the operational systems behind your sales, support, and internal ops. Inbound leads get worked in minutes, handoffs stop dropping, and AI gets wired into the steps where it actually saves hours.

Scispot
Built full automation and marketing ops stack at a B2B SaaS startup
Amex
PMO and cross-functional operations at enterprise scale
UW
Science and Business, University of Waterloo

Startup speed and enterprise process rigor do not usually come from the same person. I have worked inside both, and that is what makes the systems I build actually last.

Background in
ScispotAmerican ExpressHoffmann-La RocheUniversity of WaterlooVelocity Cornerstone
01
Your team is doing work your tools should be doing

Copying data between systems, chasing follow-ups, manually updating statuses. High-value people spending time on low-value admin because there is no system to do it for them.

02
Leads go quiet after the form submission

Someone filled out your demo request. It went into a shared inbox. Two days later a rep finally followed up. The prospect already picked a competitor. This is fixable.

03
No one knows what anything costs in time

Reporting takes three hours a week. Onboarding a new hire takes two full days of a senior person's time. Intake gets triaged by whoever happens to see it first. None of this needs to be manual.

04
Handoffs are verbal agreements that get forgotten

Work gets passed between teams through Slack messages and meeting notes. Nobody knows what is done, what is blocked, or who owns what. Things fall through not because people are careless but because the system has no memory.

05
You bought AI tools. Nothing actually changed.

Everyone has a ChatGPT account. Nobody has AI wired into the specific workflows where it would save real time. The tools are there. The integration into how the team actually works is not.

06
The process that worked at five people is breaking at fifteen

You grew fast and the operations never caught up. Now everything feels slower than it should, coordination is expensive, and fixing it keeps getting deprioritized because it is not urgent until it is.

Three services. Each ends with a working system, not a deck.

Every engagement ends with a working system: documented, tested, and handed off with 30 days of support. You get something your team can actually run from day one.

01
Lead and CRM Automation

From inbound form to qualified, followed-up, tracked lead. No manual steps, no leads going cold, no spreadsheet handoffs.

You get

A fully automated lead pipeline: form capture, CRM creation, AI enrichment, qualification routing, follow-up sequences, no-show recovery, and Slack alerts for your team. Tested and documented.

HubSpot / Airtablen8n / MakeAI enrichmentFollow-up sequencesNo-show recoverySlack alerts
02
Internal Operations Systems

Intake, escalation, approvals, and coordination built into a system with clear ownership and no lost handoffs.

You get

A structured operational workflow: centralized intake with auto-routing, SLA tracking, escalation triggers, live status visibility, and automated reporting. Built around your actual team structure.

Intake and routingSLA trackingEscalation flowsApproval chainsLive dashboardsAuto reporting
03
AI Workflow Integration

AI applied to specific parts of your workflow where it reduces real work. Scoped to the steps where it saves the most hours, not bolted on as a feature.

You get

An AI layer wired into your existing workflow: ticket triage, meeting summaries, response drafting, enrichment, or internal query routing. Scoped to where it saves the most time.

Ticket triageAI summariesResponse draftingData enrichmentInternal copilotsClaude / OpenAI API

Projects typically start at $3,500 CAD and are scoped and priced based on complexity before any work begins. No hourly billing. You know the number before you commit.

01
Scispot, B2B SaaS, life sciences
OutcomeInbound lead-to-first-contact time dropped from an average of 26 hours to under 4 minutes. No-show recovery rate improved by 40%. Zero additional headcount.

Demo pipeline automation for an inbound-led sales motion

25-person company. Sales team of 3. No dedicated RevOps. Growing inbound volume with no system to handle it.
Engagement length: 3 weeks
Similar problem? Book a call
The situation

The team was processing every inbound demo request manually. A form submission would land in a shared Gmail inbox, someone would copy the details into HubSpot when they had time, and follow-up after no-shows depended entirely on whether a rep remembered. At 20 to 30 inbound requests per week, leads were going cold before anyone reached them.

What was broken
  • No automated handoff from form to CRM: manual every time
  • Lead qualification was inconsistent across reps
  • No-show recovery was whoever remembered to follow up
  • No real-time pipeline visibility for the team
  • Slack alerts for new leads were ad hoc and often missed
The system built
  • Form submission triggers immediate HubSpot record creation with AI-enriched lead summary
  • Qualification scoring routes leads to the right rep based on company size, role, and source
  • Confirmation and prep sequences fire automatically without rep input
  • No-show detection triggers a 3-touch recovery workflow over 5 days
  • Slack alert fires within 90 seconds of form submission with full lead context
Stack
  • n8n for workflow orchestration
  • HubSpot with automated property updates and deal stage tracking
  • OpenAI API for lead enrichment and summary generation
  • Slack for team alerts
02
Enterprise operations group, financial services
OutcomeRequest-to-assignment time reduced from 2 to 3 days to under half a day. Leadership eliminated two standing status meetings per week. Escalations handled within defined SLAs for the first time.

Cross-functional intake and escalation system for an ops team at scale

Multi-team function. Cross-functional dependencies. High stakeholder coordination overhead and no standard intake process.
Engagement length: 4 weeks

Client name withheld. References available on the first call.

Similar problem? Book a call
The situation

A multi-team operations group was receiving requests through email, Slack, and ad hoc conversations. There was no structured intake, no standard escalation path, and no single view of what was in progress versus blocked. Leadership ran two 45-minute status meetings per week just to get visibility. Escalations were decided based on who was most vocal, not most urgent.

What was broken
  • Requests arrived through four channels with no triage or prioritization
  • Ownership assignment was informal and often contested
  • Escalation path did not exist: whoever was loudest got prioritized
  • Status reporting required manual compilation every week
  • Blockers had no structured path to resolution or visibility
The system built
  • Centralized intake form with auto-routing by request type and urgency tier
  • Ownership assignment on submission with SLA clock that starts immediately
  • Escalation workflow triggers automatically at 80% of SLA window, not on request
  • Live Airtable dashboard gives leadership real-time status without a meeting
  • Weekly digest auto-compiled and distributed to stakeholders every Monday at 9am
Stack
  • Airtable as operational database and tracking layer
  • n8n for routing, SLA logic, and escalation triggers
  • Slack for status alerts and escalation notifications
  • Google Workspace for digest generation and distribution
03
SaaS startup, seed stage
OutcomeMedian first-response time dropped from 6 hours to under an hour. Triage and routing eliminated as a manual task. Common issue categories now fully handled without human touchpoints.

AI-assisted support triage and first-response drafting system

Small support team of 2 handling growing inbound ticket volume. No dedicated support ops. Response times slipping as product grew.
Engagement length: 2 weeks

Client name withheld. References available on the first call.

Similar problem? Book a call
The situation

Two support reps were manually reading every inbound ticket, deciding how to categorize it, figuring out who should handle it, and writing a first response from scratch. With 60 to 80 tickets per week and two people, the math did not work. Response times were drifting, quality was inconsistent, and the team had no headroom to handle growth.

What was broken
  • No automated classification: every ticket read manually before anything moved
  • Routing was whoever saw the ticket first, not structured assignment
  • First responses written from scratch: slow and inconsistent
  • High-priority tickets had no escalation trigger
  • Common issues had no automated path: same work done repeatedly
The system built
  • Inbound tickets classified automatically by category, priority, and sentiment on arrival
  • Routing assigns to the right rep based on classification and current workload
  • Claude API drafts a context-aware first response using ticket content and the resolution knowledge base
  • High-severity tickets trigger an immediate Slack escalation before a human even opens them
  • Top 5 issue categories resolved automatically without rep involvement
Stack
  • Claude API for classification and response drafting
  • n8n for routing, escalation, and automation logic
  • Intercom as the ticketing surface
  • Notion as the resolution knowledge base the AI queries against

Those outcomes took 2 to 4 weeks. Tell me what yours would be.

Book a free 30-min call

Why this background matters for your operations.

The gap between people who know the tools and people who know the process is where most operational systems quietly fail.

I started building automation systems at Scispot, a B2B SaaS company, where I owned the full marketing and sales operations stack. That meant designing and implementing lead routing, CRM workflows, AI-assisted follow-ups, demo recovery, internal Slack alerts, and content automation, all while reporting directly on outcomes. By the end of that role, the inbound funnel ran without daily manual triage.

At American Express, I worked inside a large PMO and cross-functional operations environment. Stakeholder coordination, process governance, escalation paths, and formal handoffs at enterprise scale. The kind of complexity that breaks if you do not build it right the first time.

The combination is unusual. Startup speed meets enterprise process rigor. That is what makes the systems I build hold up past the first week.

  • Automation and Marketing Ops
    Scispot, B2B SaaS
    Startup
  • PMO and Operations
    American Express
    Enterprise
  • Pharma Technical Analyst
    Hoffmann-La Roche
    Enterprise
  • Founder's Office
    Schbang Media Services
    Agency
  • Science and Business, Co-op
    University of Waterloo
    Academic
  • Velocity Cornerstone cohort
    Waterloo Velocity, Finnav
    Founder

Four phases. One working system.

01
Operational audit (the first call)

We map the actual workflow, not the intended one. Where does work enter? Where does it stall? What is being done manually that a system should handle? This is a working session with a defined output: a clear picture of what is broken and why. The free 30-minute call is this conversation.

02
System design and scope sign-off

I design the workflow system around the specific bottlenecks we identified. You see exactly what gets built, how it will work, and what it will cost before any implementation starts. No surprises mid-project.

03
Build and test

The system is built, tested against realistic scenarios, and validated before handoff. This phase includes written documentation your team can reference, not a screen recording nobody will watch.

04
Handoff and 30-day support

You receive a system that runs. If anything breaks or needs adjustment in the first 30 days, I fix it. That is included in the project price. The goal is a system that works in production, not just in demo.

Things people ask before reaching out.

How is this different from hiring a Zapier freelancer or a general VA?

A Zapier freelancer connects two tools. That is one layer of the problem. What I build starts with understanding the operational bottleneck, designing the system around it, and then implementing across the full workflow: data routing, AI integration, escalation logic, documentation, and testing. The output is a system that keeps running, not a setup you have to maintain yourself.

We are a small team. Are we too early for this?

The best time to build the system is before a manual process costs you a customer or a hire. Most of the companies I work with are between 5 and 50 people: past the everyone-just-figures-it-out stage but not yet large enough to have a dedicated ops team. That is exactly where this work has the most leverage.

What does "30 days of support" actually mean?

It means that if anything breaks or needs adjustment in the first month after handoff, I fix it. No new scope, no new invoice. The system needs to actually work in production, not just in testing, and I stand behind that.

How long does a typical project take?

Most projects complete in 2 to 4 weeks depending on scope and how quickly we can align on requirements. The operational audit happens in week one. Design and build in weeks two and three. Testing and handoff in week four. For smaller scopes, it is faster.

Do I need to know anything technical to work with you?

No. You need to know your operations: where things are breaking and what a better outcome looks like. I handle the system design and implementation. You stay in the decisions that require your business context, and I handle the rest.

What if I am not sure whether my problem is solvable this way?

That is exactly what the first call is for. You describe the situation. I will tell you clearly whether an automated system would help, what it would take, and whether it is worth doing. If it is not the right fit, I will say so. The 30-minute call and the operational audit in week one are the same conversation.

Work with me

Tell me what is
breaking.
I will tell you if
I can fix it.

The first call is 30 minutes. You describe the operational situation. I tell you clearly whether a system can solve it and what that would look like. You leave knowing whether it is worth pursuing.

Free to book. Honest if it is not the right fit.

Background: Scispot, American Express, Hoffmann-La Roche, University of Waterloo