Customer Solutions Engineer

Overview

You're the technical expert supporting Sift's customer-facing teams (CSMs, AEs, Support). You spend half your time helping existing customers optimize their fraud detection setup—tuning ML models, analyzing why fraud slipped through, fixing integration issues. The other half is pre-sales: building custom demos for prospects, running technical discovery, proving Sift can handle their specific fraud patterns and transaction volumes.

Role Snapshot

Company Context

Stage: Late-stage (300+ employees, 700+ established customers)

Size: 301 employees

Growth: VP of CX just hired and opening roles—suggests scaling post-sales support as customer base grows

Market Position: Established fraud prevention platform with large customer base requiring ongoing technical optimization

GTM Reality

Where You Spend Time:

• Existing customers (70%): CSMs bring you in when customers have technical issues, want to optimize performance, or expand to new use cases
• Pre-sales (30%): AEs loop you into demos, technical discovery calls, POC scoping for deals over $100K
• Internal projects (10%): Building demo environments, creating technical content, training CSMs on new features

How You're Deployed:

• CSM pool: ~10-15 CSMs may request your help at any time
• Pre-sales: You might support 3-5 active deals per quarter
• You prioritize based on deal size (expansion/churn risk) and complexity
• Some weeks are heavily customer-facing (demos, workshops), others are deep technical work (analyzing fraud data, building integrations)

Internal Collaboration:

• Engineering: You escalate bugs and advocate for customer feature requests
• Product: You provide feedback on what customers actually need vs what's on roadmap
• Support: You handle escalations they can't solve (complex API issues, ML model behavior)
• Sales/CSM: You make them look smart by handling the technical heavy lifting

Competitive Landscape

Main Competitors: Other fraud prevention platforms, payment processor built-in tools, homegrown ML systems

How They Differentiate: Global fraud data network, machine learning that improves across customers, pre-built integrations vs custom build

Common Objections: "Too expensive," "Our fraud rates are already low," "We have in-house data scientists," "Implementation looks complicated"

Win Themes: Faster time-to-value than building in-house, catch fraud patterns individual companies can't see, reduce false positives vs rule-based systems

What You'll Actually Do

Time Breakdown

Customer Optimization (40%) | Pre-sales Support (30%) | Troubleshooting (15%) | Internal Work (15%)

Key Activities

• Optimization workshops with customers: Their fraud rate spiked or false positives are too high. You analyze their transaction data, review their rule configurations, recommend ML model tuning. You're basically a fraud consultant who knows Sift deeply.
• Pre-sales technical demos: Prospect says "show us how you'd catch [specific fraud pattern]." You customize the demo with their use case, walk through API integration, explain ML model training. You're selling without carrying quota.
• API integration troubleshooting: Customer's engineering team can't get Sift's API working correctly. You read their code, diagnose the issue (often data formatting or auth), write sample code, get them unstuck.
• Proof-of-concept (POC) management: Larger deals require testing Sift on their real data. You scope it, set up the environment, analyze results, build the "here's what we found" report that closes the deal.
• Fraud data analysis: Customer asks "why did this transaction get declined?" or "why did this fraud slip through?" You dig into logs, explain ML model decisions, identify gaps in their setup.
• Technical documentation and enablement: Build internal guides for CSMs ("How to diagnose false positive issues"), create customer-facing best practices, record training videos on new features.

The Honest Reality

What's Hard

• Context switching: You're pulled between a CSM's urgent customer issue, an AE's demo tomorrow, and a complex technical project. Prioritization is constant. Your calendar is fragmented.
• Customers blame the product for their config mistakes: They set up overly aggressive fraud rules, then complain Sift blocks too many orders. You tactfully explain it's their settings, then help them fix it.
• Pre-sales pressure without control: AE needs you to make the product look easy to integrate. Reality is customer's engineering team may struggle. You set expectations while staying optimistic.
• Technical depth required: You need to understand ML model behavior, API integration patterns, fraud typology, AND each customer's business. It's a wide skillset.
• Repetitive questions: Many customers ask the same things ("How does the ML model work?" "Why this false positive?"). You explain it clearly for the 100th time.
• Limited capacity: You can't support everyone at once. CSMs sometimes wait days for your help on lower-priority issues. You disappoint people by saying no.

What Success Looks Like

• Pre-sales deals you support close at 70%+ rate (vs 40-50% without SE)
• Customers you work with have better retention and lower churn risk
• CSMs trust you and bring you in proactively (not as last resort)
• Engineering team respects your technical feedback and escalations
• You reduce time-to-value for new customers (faster integrations, fewer stumbling blocks)

Who You're Working With

Internal Stakeholders:

• CSMs (your primary customers—you make them successful)
• AEs (you help them close technical deals)
• Support team (you handle their escalations)
• Product/Engineering (you're the voice of customer technical needs)

External Contacts:

• Customer fraud/risk analysts (day-to-day users asking why things work the way they do)
• Customer engineering teams (integrating Sift's API, troubleshooting issues)
• Prospect technical evaluators (CTOs, VPs Eng, fraud directors deciding if Sift can work)

What They Care About:

• CSMs: Make my customer happy, prevent churn, find expansion opportunities
• AEs: Win this deal, prove we can solve their problem, get past technical objections
• Customers: Stop fraud without blocking good customers, integrate without massive engineering effort, understand why the system does what it does
• Engineering: Is this feedback actually important? Can you reproduce the bug? What's the customer impact?

Requirements

• 3-5 years as Solutions Engineer, Sales Engineer, or Technical Support in B2B SaaS
• Strong understanding of APIs, web integrations, and how SaaS products get implemented
• Comfortable reading code (JavaScript, Python, JSON) and debugging integration issues
• Experience with fraud, security, data analytics, or ML products preferred
• Can explain complex technical concepts to non-technical audiences (and vice versa)
• Willing to occasionally travel to customer sites or conferences
• Comfortable working across time zones (EMEA and US customers both need support)