Where the work is

This role is one of those seats. You will be the technical leader for a mature, profitable B2B SaaS platform in our portfolio - an established product with real customers, real revenue, and years of accumulated code behind it. You won't advise from the outside. You'll own the codebase, set the roadmap, and be measured on the value you create.

Be clear-eyed about the shape of this job: for the foreseeable future, you are the engineering team. There is no one to delegate to and no one to hide behind. Everything technical is yours - the architecture, the deploys, the incidents, the vendor calls, the hard prioritization calls about what doesn't get built. The people who thrive in this seat are the ones who find that clarifying rather than daunting, and who create leverage through automation and good architecture instead of headcount.

The platform is established and profitable, and the mandate is to reimagine its core workflows with AI at the center, retire the expensive third-party dependencies it has outgrown, and cut the operational drag that currently taxes engineering. You're the technical leader who owns that from the front.

We're looking for a jack of all trades, a strong generalist who can go deep where it counts. The good news on the stack: this business is already close to where Beacon is going - Python and TypeScript services on AWS, React on the front end, Kubernetes, Postgres, Terraform. There is no legacy-platform tax to pay here. The leverage is in delivery discipline, in the parts of the platform that grew organically and now need real architecture, and in the AI surface that doesn't exist yet. You'll be the one who builds all three.

What you'll do

• Own the technical and product roadmap. Decide what gets built and why, in partnership with the GM, the pod's product team, and customers. Translate business priorities into a sequenced engineering plan you're accountable for - and defend the tradeoffs when everything is urgent.
• Drive everything technical. Architecture, code quality, security, data modeling, performance, release engineering, and the migration path off the vendors we no longer want to depend on. You set the bar and you make the final technical decision.
• Ship AI that changes the business, not demos. Take the AI surface from plan to production: evals, guardrails, tenant-safe tool use, and features customers feel. You know when to lean on a foundation model, when a deterministic pipeline is the correct answer, and how to keep agentic workflows on the rails in a multi-tenant system.
• Operate as a team of one, and build like you won't be. Ruthless prioritization, small safe changes, and automation over heroics. Leave behind a codebase, a test suite, and documentation that a second and third engineer can join without a three-month archaeology project - because if you create the value, that's the next conversation.
• Increase velocity and harden the codebase. Trunk-based development, real CI, a staging environment on synthetic data rather than stale production records, feature flags, meaningful test coverage, and a deploy path that doesn't depend on you remembering it. Then pay down the risk in the parts that grew organically so the platform can carry far more weight than it does today.
• Own the operating cost. You have enough cloud and vendor cost literacy to find the line item nobody has looked at, and the engineering judgment to know whether the fix is worth the week.
• Own security and data handling as engineering work. Least-privilege access to production, sensitive data kept out of non-production environments, and a posture you can defend to Beacon Legal and to a customer's security review.
• Be the technical face of the business. Earn the trust of operators and customers, set expectations honestly, and communicate clearly in writing.

What we're looking for

• A senior Python and TypeScript engineer - day one. Deep, current experience with Python backend services and Node/TypeScript, comfortable owning a large, mature codebase immediately: reading it, fixing it, and reshaping it with judgment. This is the non-negotiable starting requirement.
• A genuine generalist ("jack of all trades"). You're T-shaped and you pick up new stacks quickly and enjoy it. You can move across web, mobile, data, infra, and AI without waiting for permission.
• Cloud and SQL depth. You've designed, deployed, and operated real systems on AWS - containers on Kubernetes, queues and event-driven workers, infrastructure-as-code - and you can model, query, tune, and reason about production Postgres.
• Release engineering conviction. You have personally introduced trunk-based development, automated testing, and environment discipline to a codebase that didn't have them, without stopping delivery to do it.
• AI-native engineering. You've shipped production software on foundation models - retrieval, tool use, evals, guardrails - not just prototypes, and you have a point of view on doing it safely in a multi-tenant product.
• Proven sole ownership. You've been the only engineer, or the clear technical owner, on a system that mattered: design, build, deploy, on-call, and the customer conversation when it broke. You know how to carry that without burning out or cutting the corners that come back.
• Product and business instinct. You can read what actually drives the business, prioritize accordingly, and defend a roadmap to a GM and to customers. You can tell which technical debt is actively costing money and which is merely ugly, and you spend your credibility on the former.
• A strong writer and communicator who builds trust with non-engineers. In a team of one, written clarity is how the rest of the business knows what's happening.

Nice to have

• Multi-tenant SaaS architecture, and the isolation and access-control problems that come with it.
• Event-driven and queue-backed systems at meaningful volume - throughput, backpressure, idempotency, and what happens when a downstream provider degrades.
• Cross-platform mobile experience, and comfort maintaining a mobile release channel alongside everything else.
• Third-party integration work at depth: REST APIs, webhooks, file-based handoffs, and the unglamorous reality of integrating with customer systems that have no API and no documentation.
• Security and privacy engineering - SOC 2, PII minimization, encryption, secrets management, audit logging.
• Experience retiring a costly vendor dependency by building the replacement in-house, without a disruptive rewrite.
• Vertical or B2B SaaS experience, especially where the buyer and the end user are different people.