Full-Stack Software Engineer, AI-Enabled Platform at Directive – Canada | Jobyssey

Directive Consulting is the leading B2B marketing agency for brands that are serious about the metrics that matter most.

With over a decade of expertise and 420+ B2B brands under our belt, we've earned our reputation on real outcomes. We run through 3 divisions: Performance, Communications, and Commerce, all wired to share intelligence and answer to revenue. That system spans everything from Paid Media and Content Marketing, to PR and Paid Social, to Marketplace and Lifecycle Marketing. Tying it all together is our methodology, DiscoverabilityOS™, which aligns brand and demand to guide your ICP to choose you across every decision point.

We move fast. We think boldly. We hire people who challenge the status quo, build campaigns worth talking about, and make bold brands impossible to ignore. If you live for performance, love solving hard problems, and want your work to actually mean something, you'll fit right in.

The Role

We're looking for a Full-Stack Software Engineer to join our Operational Technology team and take internal ownership of the web layer of Stratos, Directive's AI-powered marketing analytics platform, along with our company website and other internal web surfaces. Stratos has been built in partnership with an external agency, and we're bringing the web engineering in-house to move faster, respond to our internal team and clients in hours instead of weeks, and compound platform knowledge that stays with Directive. You'll be the first dedicated internal owner of this surface, with a clear growth path to senior scope as Stratos does. AI-assisted development is core to how you work: you use tools like Claude Code, Cursor, and Copilot to move from spec to production fast, while still shipping clean, reviewed backend services and React/TypeScript that we can maintain for years.

Roles & Responsibilities

Own and evolve the web surfaces that power Directive: the internal strategist portal, the external client portal, the reporting builder, campaign management tools, admin and portfolio surfaces, and the company marketing site
Build and maintain the ads and campaign management UI on top of third-party ad APIs, including campaign builders, bulk edit flows, audience and budget management, and creative tooling
Build and maintain data metrics views and dashboards backed by BigQuery, MySQL and Postgres including widget components, chart rendering, and the queries that drive them
Manage containerization via Docker and AWS infrastructure, overseeing EC2 deployments and DevOps pipelines with a focus on reliable delivery and automation.
Manage GitHub Actions and automated CI/CD pipelines; ensures reliable iteration within staging and seamless delivery to production
Design and optimize relational database schemas and data structures
Maintain highly-available and scalable architectures; oversees Nginx setups, load balancing, and secure infrastructure to ensure system resilience
Own the reporting UI end-to-end: report list, create, details, template management, PDF export, drag-and-drop widget reorder, and widget-level date overrides
Own the admin and portfolio surfaces: client CRUD, user management, OAuth flows, CRM integration configuration, and alerting
Own the background and integration layer: queue-based sync jobs, CRM sync (Salesforce, HubSpot), ad platform integrations (LinkedIn, Google Ads, Meta), data pipelines, auth and sessions, logging, and error tracking
Partner on the AI-facing UI layer: surface LLM outputs, streaming responses, tool runs, and agent conversations cleanly, with clear states for loading, error, and partial results
Use AI coding tools (Claude Code, Cursor, Copilot) as a daily accelerator to scaffold, iterate, and refactor production code, with strong review instincts so AI-generated code meets our bar
Translate Figma mockups and loose specs into production-ready components, owning design system usage, accessibility (WCAG), and performance
Handle auth, RBAC, and safe data exposure patterns (row-level security, parameterized queries, least-privilege keys) on the web tier
Add telemetry and error tracking so issues are detected and diagnosed before clients see them, and product decisions are backed by data
Absorb ownership of modules from our outgoing agency partner: review existing code, document patterns, stabilize rough edges, and evolve them inside our internal roadmap
Ship, support, and iterate: fix bugs, improve performance, reduce load times, and evolve UX based on real usage
Document components, patterns, and trade-offs so the team can build on top of your work

What You Offer /Qualifications

Experience: 5+ years of professional software engineering with production ownership of user-facing applications.
Backend: Production Node.js required. This is a hard requirement, not a nice-to-have. API design, background jobs, auth flows, and the database access layer should be work you've owned end-to-end, not just supported.
Frontend: React and TypeScript are table stakes for this role. We're evaluating backend depth and system ownership beyond that baseline. Comfortable with modern frameworks (Next.js, Remix, or similar); fluent in component design, state management, and performance profiling
Full-Stack Range: Owns the backend glue as well as the UI: API routes, serverless functions, auth flows, database queries, background jobs
Cloud Delivery: AWS familiarity strongly preferred — EC2, Lambda, SQS, S3, or equivalent. Ships to managed platforms with CI/CD and observability in place.
AI-Assisted Development: Uses Claude Code, Cursor, Copilot, or equivalent as a daily tool to accelerate real engineering. You ship production code, not platform-locked low-code output. You know how to prompt, review, and refactor AI-generated code without shipping AI-generated bugs
LLM Integration: Experience wiring LLM APIs (Anthropic, OpenAI, Google) into web apps: streaming, tool/function calling, structured outputs, timeouts, retries, guardrails
Data & SQL: Comfortable querying MySQL, Postgres and BigQuery; can shape data for UI consumption and validate results
Advanced SQL & Query Layer: Mastery of Knex and MySQL to architect high-performance, scalable data structures and queries
Relational & Real-time Data: Proficient in MySQL, Postgres, and Supabase (or equivalent); deeply understands RLS, storage, edge functions, and real-time subscription patterns
APIs & Integrations: REST/GraphQL, OAuth2/OIDC/JWT, webhooks, pagination, rate limits, idempotency; comfortable integrating with HubSpot, Salesforce, LinkedIn Ads, Google Ads, GA4, or equivalent
Background Jobs: Familiar with queue-based systems (BullMQ, SQS, or similar) for sync jobs, scheduled tasks, and long-running work
Design Sensibility: Takes a Figma file or a rough sketch and produces a polished, on-brand, accessible UI without a designer holding your hand
Ownership & Communication: Operates well with imperfect specs and inherited code. Asks the right questions, ships iteratively, and drives outcomes

Additional Things That Will Impress Us:

Design systems: shadcn/ui, Radix, Tailwind, or similar; building and maintaining shared component libraries
Marketing data literacy: CRM (Salesforce/HubSpot), GA4, paid media APIs (LinkedIn, Google Ads, Meta), attribution and measurement basics
Security & compliance: least-privilege IAM, secrets management, PII handling, SOC 2/ISO 27001 awareness
Performance: Core Web Vitals, bundle analysis, caching strategies, edge rendering
Experimentation: feature flags, A/B exposure, analytics instrumentation habits that support decision-making
Reporting UX: experience with dashboarding frameworks, PDF generation, chart libraries (Recharts, Tremor, Visx), drag-and-drop builders
Codebase inheritance: prior experience absorbing a production codebase from an outgoing team and stabilizing it

What Success Looks Like

Within 30 days: ships fixes on the inherited bug backlog and keeps shared web surfaces healthy, with clear handoff notes from the outgoing agency captured in writing
Within 90 days: ships at least one production web surface end-to-end (reporting module, portal page, or admin flow) with telemetry, error tracking, and a clear rollout plan
Within two quarters: absorbs meaningful ownership of inherited Stratos modules, with stable on-call coverage and a documented understanding of the background queue and integration layer
By end of year one: takes ownership of net-new Stratos features outside the scope of the outgoing agency, measurably compressing internal bug response time and reclaiming engineering capacity that had been pulled into maintenance
Raises the web engineering bar: contributes reusable components, patterns, and AI-assisted workflows that peers adopt across Stratos and internal tools

Level-Specific Expectations

At Mid-Level:

Implements designs end-to-end with clean code, tests, and docs
Uses AI tools effectively to move fast without sacrificing quality
Ships bug fixes and UI improvements on inherited modules from day one
Raises risks early and owns staged rollouts of the work they lead
Collaborates cleanly across engineering, product, and ops

At Senior:

Leads the design of cross-team web products and defines component patterns, auth and data-exposure guardrails, and delivery cadence
Proactively manages performance, accessibility, and reliability across the web surface
Mentors peers through code reviews and pairing
Owns the relationship with any remaining external contractors or specialist vendors

Explore more jobs