ManManager, Software Engineering - Agentic Assurance Platform

About Invoca

Invoca is an AI-powered revenue execution platform that brings together marketing, commerce, and contact center teams to turn every customer interaction into measurable, profitable growth. Join our dynamic, fast-growing team, where innovation and collaboration are at the core of our culture.

About the Team

Invoca's AI agents’ outcomes depend on much more than the model and harness: tool calls, partner APIs, configuration, telephony timing, agent-to-agent collaboration, context modeling, and data delivery all have to work together, and any one of them can change on any given day. In agentic systems, the hardest failures are quiet ones: an agent can report success while the booking, the lead, or the data never arrives.

The Agentic Assurance Platform team builds the product that gives Invoca, and its customers, justified confidence that every agent deployment still works after something changes. It is a platform, not a QA function: a shared system that turns any change, on any surface, into a verdict backed by evidence. In practice that means APIs and a policy engine for release gates in CI, orchestration of simulated conversations across voice, messaging, web, agents, scorers and LLM judges calibrated against human review, live quality signals on production traffic, and checks that verify outcomes where they land, in the systems customers rely on.

The team is new and moving fast, and its charter is broad: voice and messaging agents, and every agentic product Invoca ships as that portfolio grows.

About the Role

We're hiring an Engineering Manager to lead the Agentic Assurance Platform team. You'll manage a growing team of engineers and own the platform's delivery, operational health, and growth, reporting to AI Engineering leadership. You'll join a new team with a clear mission and the chance to shape both the product and the team. Your job is to build assurance into a platform that every agent team at Invoca adopts by default: every agent-affecting change checked before it reaches customers, every live agent monitored, and every failure turned into a test so it cannot ship twice.

This is a management role with technical teeth. You'll work alongside an architect and a tech lead, so you won't be the primary author of the architecture, but you need the depth to hold a high bar in design reviews of distributed systems, APIs, data pipelines, and AI evaluation, and to earn the trust of senior engineers. Assurance adds hard calls of its own, such as what counts as evidence, when a sample is too small, and whether a judge can be trusted, so you'll need to stay close to the code, the traces, and the data.

What You'll Do

Lead and grow the team

• Manage, coach, and develop the engineers on the Agentic Assurance Platform team: clear expectations, honest feedback, real career growth.

• Hire senior talent in a competitive market, and own candidate experience that top-tier passive candidates say yes to: tight loops, prepared interviewers, fast decisions.

• Build a culture of ownership and craft, where engineers build for their users, measure what they ship, and use the agentic tools they build.

Own platform outcomes, not roadmap items

• Run the platform as a product with internal users: set adoption and outcome goals (failures caught by automation first, time to detect, time to restore), instrument them, and measure whether teams ship faster and with more confidence because of your team's work.

• Partner with Product, the agent platform and product teams, the teams that deploy agents for customers, and architects on roadmap and priorities, and represent the platform in R&D planning and launch readiness.

• Make the platform self-serve: APIs, SDKs, CI integrations, shared scenario and policy formats, and reference implementations that let other teams add their own checks without waiting on yours.

Build the platform as a product

• Build a platform that governs every way an agent's behavior can change, not only code: prompts, templates, model routes, tool contracts, configuration, and partner systems. Ship new gates in shadow mode, and earn the right to block.

• Design for risk-based policy: a copy edit and a model swap get different evidence, sample sizes, and rollout, decided by policy the platform enforces rather than by a review meeting.

• Keep the core small and stable: it owns contracts, policy, and evidence, while simulation, scoring, and observability engines plug in behind adapters, so swapping an engine never changes a verdict.

Run delivery with data

• Own delivery health using DORA-style metrics and flow analysis, and manage flow and work in progress, not just effort.

• Make delivery predictable: honest estimates, visible risks, no surprise slips.

• Find where high-volume, low-difficulty toil concentrates, and automate it.

Build trust into the platform

• Treat judges, simulators, test fixtures, monitors, and the gate itself as instruments, with calibration and health checks built into the platform. Missing evidence means "insufficient," never "pass."

• Make the platform's numbers trustworthy: rates with confidence intervals, judge accuracy against human labels, and open doubts recorded alongside every verdict.

• Make production-like testing a platform capability: isolated test environments on demand and realistic simulations that keep test traffic separate from customer data.

• Embed operational rigor: platform SLOs, alerts that reach an owner, on-call health, honest postmortems, and a pipeline that turns every incident into a regression scenario. Partner with Security so access control, audit, and privacy keep pace with capability.

Champion the agentic development era

• Make your own team the reference implementation of AI-assisted engineering: agentic coding workflows, automated review gates, AI in the SDLC as habit rather than demo.

• Put agents to work inside the platform itself, generating scenarios, triaging failures, and finding coverage gaps, under the same gates as everything else.

What We're Looking For

• 8+ years in software engineering, including meaningful hands-on years building backend, distributed, data, or platform systems, and 3+ years managing engineers.

• Experience leading a platform, infrastructure, or developer-experience team whose customers are internal engineers, with a track record of driving measurable adoption. Experience building evaluation, testing, or reliability infrastructure as a shared platform is a strong plus.

• Working fluency in modern AI systems and how to evaluate them: LLM APIs and agents, tool calling with MCP or similar protocols, simulation, LLM-as-judge and its failure modes, tracing and telemetry. Enough depth to hold a technical bar in design reviews; you don't need to be a researcher.

• Statistical judgment: comfortable with sampling, confidence intervals, and comparing a candidate against what's live, and able to explain why 30 passing runs isn't proof.

• Systems thinking across boundaries: you've debugged failures that span services, third-party APIs, data pipelines, and configuration, and you know most production failures happen at the seams.

• Data-driven delivery management: you've used DORA metrics or equivalent to find and fix real bottlenecks, and you understand flow.

• Operational and security mindset: you've owned production services, run incident response, and treat auditability and the privacy of customer conversations as design requirements.

• Demonstrated success hiring and retaining senior engineers.

• Direct, calibrated communication with engineers and executives alike: honest characterization over spin, risks surfaced early, and clarity about what is known and what isn't.

• High ownership in ambiguous, fast-moving territory.

• A plus: experience with voice or conversational AI, telephony or contact-center systems, or regulated industries such as healthcare or financial services.

How Success Will Be Measured

Success in this role will look like:

• The assurance platform is the default way agent teams ship: agent-affecting changes run through it, teams add their own checks self-serve, and platform telemetry, not anecdote, proves it.

• Agent problems are caught by automation (a gate, a simulation, a live eval, or an alert) before they reach customers, and every agent-affecting change carries a recorded verdict, whatever surface it comes from.

• Failures are detected fast, and each confirmed failure becomes a regression scenario within days, without a manual hand-off.

• Every go-live is proven ready in isolated environments, and the quality numbers leadership and customers see are calibrated, with their uncertainty stated.

• Delivery is predictable and measurably improving.

• The team is retained, growing, and hiring well.

📍 Location: This is a remote-first role. We are currently hiring in the following locations: 📍

United States: Greater Los Angeles Area (including Santa Barbara and San Diego) · SF Bay Area · Denver Metro · Austin Metro · Chicago Metro · Greater NYC Area · Seattle Metro

Candidates must be based within ~2 hour drive of these areas. Occasional business travel may be required.

Compensation, Benefits & Perks:

At Invoca, we believe great work starts with taking care of our people. That's why we offer competitive compensation and benefits designed to support your health, your family, your growth, and your life outside of work. For teammates in the U.S., health benefits begin on your first day of employment.

Compensation

• Base Salary: $175,000–$275,000*

• Variable Compensation: Bonus and/or commission eligibility, where applicable.

• Equity: All employees are invited to share in Invoca's success through our stock option program.

*Actual compensation will depend on factors including experience, skills, and location.

Benefits & Perks

Health & Wellbeing

• Day-One Health Benefits – Medical, dental, and vision coverage begins on your first day of employment for U.S.-based teammates.

• Mental Wellbeing – Access to mental wellbeing support and an Employee Assistance Program (EAP).

• Wellness Subsidy – Reimbursement that can be applied toward gym memberships, fitness classes, and more.

Time Away

• Flexible Time Off – We encourage a healthy work-life balance. Our flexible paid time off policy gives you the flexibility to recharge and take time away when you need it.

• Paid Holidays – Invoca provides 20 U.S. paid holidays, including a winter break, giving you time to rest, recharge, and spend time with friends and family.

• Paid Family Leave – Up to 12 weeks of 100% paid leave for baby bonding, adoption, and caring for family members.*

• Paid Medical Leave – Up to 12 weeks of 100% paid leave for childbirth and medical needs.*

Growth & Career

• Enterprise-Grade AI Tools – Access to industry-leading enterprise AI tools, including leading LLMs and agentic AI workflow solutions to help you maximize productivity and work more effectively.

• Professional Development – Annual reimbursement of up to $1,000 per fiscal year to support learning, certifications, conferences, and other professional development opportunities.

Financial Benefits

• 401(k) – Invoca offers a 401(k) plan through Fidelity with a company match of up to 4%.*

Recognition & Community

• Recognition Programs – We celebrate great work through company-wide and team recognition programs, including President's Club and quarterly and departmental awards recognizing innovation, execution, and customer impact.

• InVacation (Sabbatical) – As a thank-you to our long-term team members, we offer a sabbatical after seven years of service.

*Some benefits may vary for teammates outside the U.S. based on local laws, regulations, and country-specific offerings.