Staff Technical Product Manager, Internal AI Applications
•Unity· 2h
📦 Product⏰ Full-time🎯 Lead / Principal💰 $153k-230k🇨🇦 Canada
2h ago
Oscilar·6 months ago
Shape the future of trust in the age of AI
At Oscilar, we're building the most advanced AI Risk Decisioning™ Platform. Banks, fintechs, and digitally native organizations rely on us to manage their fraud, credit, and compliance risk with the power of AI. If you're passionate about solving complex problems and making the internet safer for everyone, this is your place.
Mission-driven teams: Work alongside industry veterans from Meta, Uber, Citi, and Confluent, all united by a shared goal to make the digital world safer.
Ownership and impact: We believe in extreme ownership. You'll be empowered to take responsibility, move fast, and make decisions that drive our mission forward.
Innovate at the cutting edge: Your work will shape how modern finance detects fraud and manages risk.
We are seeking an innovative AI Product Manager to lead the development of our next-generation risk orchestration platform. You will be building the future of risk decisioning - a visual workflow automation platform meets cutting-edge AI, specifically designed for financial risk workflows. This platform will combine no-code visual workflow builders with Generative AI and Agentic AI capabilities to automate complex real-time decisioning for fraud, credit, onboarding, and AML use cases. This is a unique opportunity to define how financial institutions will orchestrate risk operations in the AI era.
You'll create a sophisticated visual workflow builder specifically designed for financial risk operations:
Drag-and-Drop Risk Nodes: Pre-built components for common risk operations (KYC checks, transaction monitoring, credit scoring, sanctions screening) that connect seamlessly
Real-Time Execution Canvas: Visual feedback showing live data flow, with ability to inspect outputs at each node and test individual steps without running entire workflows
AI-Powered Decision Trees: Visual branching logic that adapts based on risk signals, with AI suggesting optimal paths and thresholds
Risk-Specific Templates: Pre-built workflows for common scenarios like "High-Risk Transaction Investigation" or "Enhanced Due Diligence Flow"
Going beyond basic LLM nodes, you'll productize advanced AI capabilities:
Agentic Risk Investigators: Autonomous AI agents that can navigate multiple data sources, conduct investigations, and escalate findings - think of nodes that don't just process data but actively investigate
GenAI Document Analysis: Nodes that extract insights from unstructured documents (bank statements, incorporation docs, adverse media)
Adaptive Learning Nodes: Components that improve their decisioning based on feedback loops and case outcomes
Multi-Model Orchestration: Visual ensemble methods where multiple AI models vote on decisions with configurable weighting
Sub-100ms Performance: Real-time execution handling millions of decisions daily with queue management and worker distribution
Error Recovery Workflows: Visual error handling with automatic retry logic and fallback paths for failed services
Compliance Recording: Every decision path logged with full explainability for regulatory review
Version Control Integration: Git-based workflow management with staging/production environments
Define the product vision for an AI-first risk orchestration platform that revolutionizes how FIs manage risk workflows
Develop the roadmap that balances no-code accessibility with sophisticated AI capabilities
Research and analyze competitive landscape (n8n, Zapier, Temporal) to identify differentiation opportunities
Create the strategy for transitioning from traditional rule engines to AI-orchestrated decisioning
Define success metrics and KPIs that demonstrate value across fraud, credit, AML, and onboarding use cases
Agentic Workflow Components: Design autonomous AI nodes that can:
Conduct multi-step investigations across data sources
Generate SAR narratives with supporting evidence
Perform iterative searches until finding conclusive results
Make decisions and explain reasoning in natural language
Learn from investigator feedback to improve future decisions
LLM Integration Framework: Build a flexible system supporting:
Multiple LLM providers (OpenAI, Anthropic, local models)
RAG nodes for policy and regulation lookups
Prompt engineering interface within nodes
Token usage monitoring and cost optimization
Model performance A/B testing within workflows
Intelligent Orchestration: Create smart workflow features:
AI-suggested next nodes based on current data flow
Automatic workflow optimization (identifying bottlenecks)
Anomaly detection on workflow performance
Dynamic routing based on confidence scores
Self-healing workflows that adapt to API changes
Human-in-the-Loop AI: Design collaborative features:
AI draft decisions with human approval nodes
Feedback collection nodes that train models
Explainability widgets showing AI reasoning
Confidence thresholds triggering human review
Audit trails for all AI decisions
Node Library Development: Design 200+ pre-built risk nodes including:
Data ingestion nodes (core banking APIs, bureau connectors, blockchain analyzers)
Enrichment nodes (identity verification, device fingerprinting, graph analytics)
Decision nodes (ML scoring, rules engines, policy tables)
Action nodes (case creation, alert generation, automated responses)
AI nodes (investigation agents, narrative generators, anomaly detectors)
Workflow Studio Features: Create an intuitive builder with:
Split-screen design showing workflow logic and live data flow
Node search with AI-powered recommendations ("nodes similar to velocity checks")
Sticky notes and documentation directly on the canvas
Workflow simulation with synthetic data for testing
Performance profiling showing latency at each node
Template Marketplace: Curate risk-specific workflows:
"FATF-Compliant Transaction Monitoring" template
"SaaS Fraud Prevention Suite" with pre-configured rules
"Instant KYC + Credit Decision" for lending
Community-contributed workflows with security vetting
Developer Experience: Enable power users with:
JavaScript/Python code nodes for custom logic
Webhook nodes for real-time event triggering
API nodes with OAuth management
Git sync for workflow version control
CLI tools for workflow deployment
Collaborate with engineering to build a platform that's both powerful and performant
Work with risk domain experts (fraud, AML, credit) to ensure workflows meet practitioner needs
Partner with customer success to gather feedback and iterate on platform capabilities
Coordinate with sales and marketing to articulate the platform's unique value proposition
Lead agile ceremonies and maintain clear communication across all stakeholders
5+ years of product management experience, with at least 3 years in B2B SaaS platforms
Proven track record building workflow automation, orchestration, or integration platforms
Experience with AI/ML products, particularly in production environments
Background in fintech, risk management, or financial services preferred
Demonstrated success launching technical products for non-technical users
Deep understanding of workflow orchestration patterns, event-driven architectures, and state machines
Experience with visual programming concepts and node-based editors (Node-RED, Retool Workflows, Pipedream)
Knowledge of AI/ML concepts including LLMs, embeddings, vector databases, RAG, and agent architectures
Understanding of real-time systems, message queuing (Kafka, RabbitMQ), and streaming data processing
Familiarity with API design, webhooks, OAuth flows, and integration patterns
Experience with workflow execution engines (Temporal, Airflow, Camunda)
Understanding of no-code/low-code platforms and visual programming paradigms
Understanding of risk decisioning workflows across fraud, AML, credit, and onboarding
Knowledge of regulatory requirements for automated decision-making (FCRA, GDPR Article 22)
Familiarity with risk operations including case management and investigation workflows
Understanding of financial services infrastructure and data sources
Strong analytical skills with ability to translate complex technical concepts into user value
Excellent written and verbal communication skills
Data-driven decision making with experience in A/B testing and experimentation
User research and design thinking capabilities
Ability to balance technical innovation with practical business needs
Technical background in Computer Science or Engineering
Experience with specific technologies: Node-RED, Apache Airflow, Temporal, or similar workflow engines
Knowledge of specific risk platforms (Actimize, SAS, Feedzai, etc.)
Experience building developer platforms or tools
Understanding of MLOps and model deployment pipelines
Background in building real-time, mission-critical systems
Compensation: Competitive salary and equity packages, including a 401k plan
Flexibility: Remote-first culture — work from anywhere
Health: 100% Employer covered comprehensive health, dental, and vision insurance with a top tier plan for you and your dependents (US)
Balance: Unlimited PTO policy
Technical: AI First company; both Co-Founders are engineers at heart; and over 50% of the company is Engineering and Product
Culture: Family-Friendly environment; Regular team events and offsites
Development: Unparalleled learning and professional development opportunities
Impact: Making the internet safer by protecting online transactions