Overview
You'll build and maintain the GTM infrastructure for Roboflow's sales team selling computer vision developer tools. This means Salesforce administration, pipeline reporting, forecasting models, and trying to enforce data hygiene with a team that's growing fast. You work directly under Ashley (Head of RevOps) on a small, growing team.
Role Snapshot
| Aspect | Details |
|---|---|
| Role Type | GTM Systems & Operations |
| Sales Motion | Product-led growth + sales-assisted (developers convert to enterprise) |
| Deal Complexity | Mix - PLG conversions are transactional, enterprise deals are consultative |
| Sales Cycle | PLG: Days to weeks / Enterprise: 2-4 months |
| Deal Size | $5K-$100K+ ACV (wide range) |
| Quota (est.) | N/A (supporting quota-carrying reps) |
Company Context
Stage: Series B (recently raised, exact amount not disclosed)
Size: ~50-100 employees (growing rapidly based on hiring signals)
Growth: Expanding NYC office, active hiring across GTM, strong momentum in 2025
Market Position: Category leader in computer vision tools for developers - competing with build-your-own solutions and horizontal ML platforms
GTM Reality
Pipeline Sources:
- 60-70% PLG - developers sign up for free tools, use product, then convert to paid or get flagged for sales outreach
- 20-30% Outbound - targeting ML teams at companies building computer vision applications
- 10% Partnerships/Referrals - integrations with cloud providers, hardware vendors
SDR/AE Structure: Likely hybrid model - AEs self-source some PLG leads, SDRs/BDRs handle outbound to enterprise accounts
SE Support: Product is technical enough they probably have solutions engineers or technical AMs for complex POCs
Competitive Landscape
Main Competitors: Google Cloud Vision AI, AWS Rekognition, custom-built ML pipelines, other computer vision platforms
How They Differentiate: Developer-first tooling, easier annotation/training workflows, better model deployment than building from scratch
Common Objections: "We can build this internally", pricing vs DIY approach, data security/compliance for enterprise
Win Themes: Speed to production, better developer experience, superior annotation tools, active community
What You'll Actually Do
Time Breakdown
Salesforce Admin (35%) | Reporting/Analytics (30%) | Process/Enablement (20%) | Cross-functional Projects (15%)
Key Activities
- Salesforce Hygiene: You're constantly cleaning data, fixing duplicate records, updating fields that reps didn't fill out, and building validation rules that people will find workarounds for. This never ends.
- Pipeline Reporting: Building and maintaining weekly/monthly dashboards for leadership. Forecast accuracy reports, conversion metrics by stage, rep performance tracking. You'll spend hours reconciling why the numbers don't match between systems.
- Tool Integration: Managing the stack - Salesforce, product analytics (probably Amplitude/Mixpanel), outreach tools, enrichment (ZoomInfo/Clearbit), maybe Gong. Fixing broken syncs, managing licenses, fielding "why isn't this working" Slack messages.
- Process Documentation: Writing playbooks that ideally reps would follow - lead routing rules, opportunity stage criteria, forecasting methodology. Realistically, you'll spend more time chasing people to use the process than building it.
- Cross-functional Alignment: Translating between sales ("we need more leads"), marketing ("we're sending qualified leads"), and product ("usage data shows different story"). Lots of meetings to align on definitions and metrics.
The Honest Reality
What's Hard
- Data Quality Is A Losing Battle: In a PLG motion, you have self-serve signups, product usage data, and manual sales entries all flowing into CRM. Things break constantly. Reps don't update stages. You're always 80% clean at best.
- Reporting Complexity: Leadership wants to know "what's our pipeline from PLG vs outbound" but the attribution is messy. Developers sign up free, ghost for 3 months, then come back via sales outreach. What source gets credit? You'll build the same report 5 different ways.
- Fast Growth = Constant Change: Processes you build this quarter are outdated next quarter. They're hiring quickly, adding new segments, changing comp plans. You're rebuilding systems while people are using them.
- RevOps Is Reactive: Your roadmap gets interrupted daily with "can you pull this report for exec team by EOD" or "Salesforce is broken, help." Strategic projects get pushed for urgent firefighting.
What Success Looks Like
- Forecast accuracy within 10% at month/quarter end
- Clean pipeline reporting that leadership trusts for board meetings
- Reduced time-to-productivity for new sales hires (they can navigate systems and find what they need)
- Data hygiene metrics improving quarter-over-quarter (even if never perfect)
Who You're Selling To
Primary Buyers:
- ML Engineers / Data Scientists (users of the product)
- Engineering/ML Managers (budget holders for team tools)
- VP Engineering / CTO (enterprise deals)
What They Care About:
- Speed to deploy models vs building infrastructure from scratch
- Ease of annotation and data labeling workflow
- Model accuracy and performance benchmarks
- Integration with existing ML stack (cloud providers, frameworks)
- Cost vs build-your-own (both money and engineering time)
Requirements
- 2-4 years experience in RevOps, Sales Ops, or similar GTM operations role
- Strong Salesforce administration skills (flows, validation rules, reports/dashboards)
- Experience with PLG or developer-focused SaaS (understanding product-usage data, self-serve funnels)
- SQL or similar data querying ability to pull reports outside of standard tools
- Comfortable working in ambiguity - processes aren't mature, you'll be building things from scratch
- Detail-oriented but pragmatic - can't let perfect be enemy of done in a fast-growth environment