Senior Software Engineer - AI Lead - Asset Management

Beyond Potentials; Le Vaud; On-site; CHF 140,000 - 190,000; Full time

Job summary

Beyond Potentials seeks a Senior Software Engineer - AI Lead to drive production-grade software on our data platform, leveraging AI capabilities. You will collaborate with platform, data, and portfolio teams to ship high‑quality features, enforce CI/CD, and ensure robust security across the pipeline.

Applicants will have 8+ years in production software, strong Python and a statically typed language, plus experience with Kubernetes, IaC, and AI tooling.

Qualifications

• 8+ years building production software, with strong Python and statically typed languages.
• Experience in cloud, Kubernetes, and CI/CD pipelines; IaC mindset.
• Proven ability to design APIs, ensure security, and manage data-friendly AI components.

Responsibilities

• Ship features from concept to production with a small, high‑quality team.
• Set code standards: typing, linting, contract tests, pinned dependencies.
• Own delivery: IaC, CI/CD, container builds, environments, releases and rollback.
• Build security into the pipeline: secrets, least‑privilege IAM, scanning, SBOM.
• Assemble AI components: LLM gateway, access control, prompts and versioning.
• Make evaluations a release gate and run monitoring, tracing, alerts.
• Review and pair with junior engineers; enforce CI‑driven quality.

Skills

Python

Go/Java/TypeScript

Kubernetes

Terraform

CI/CD

API design

Security

AI/LLM

Tools

CDK

Docker

Job description

We have been mandated exclusively by one of our client, an important Asset Manager, for a Senior Software Engineer Role – AI Lead. In a dynamic team, they are building the future of the company ain terms of Technology and what they build sits on their data platform and increasingly uses AI. In terms of daily business, you will bring what keeps software alive after the demo: design, tests, observability, security, and the anti-patterns you have already paid for. The person we are looking for is an Hands-on senior engineer who can work with the platform engineer, the data platform, and the portfolio managers and analysts who write their own code.

In terms of description of what will be the position, here are some details of what you will need to do:

• Ship. Work with portfolio managers, analysts and operations to take use cases to production. Small scope, production quality.
• Set the code standard. Strict type checking and linting as CI gates, contract tests at service boundaries, pinned dependencies with automated updates, mandatory review, decision records for choices that are hard to reverse.
• Own delivery for what the team ships. Infrastructure as code, CI/CD, container builds, environments, releases and rollback. The platform engineer owns accounts, clusters and networking; you own everything from the container image up.
• Build security into the pipeline. Secrets management, least-privilege IAM roles for the services you ship, dependency and code scanning, SBOM and signed artefacts, policy as code. Four-eyes approval on every production change, enforced by the pipeline. Changes to pipelines, policies and IAM reviewed by the security team; repository and CI admin rights sit outside the engineering team; production access is break-glass only, logged and reviewed after use.
• Assemble the shared AI components, adopting before building: LLM gateway (model access, data residency, logging, cost, evaluation hooks); retrieval with access control enforced at query time; agent runtime with tool calling and human-in-the-loop validation; prompt and configuration versioning.
• Make evaluation a release gate. Build and run evaluations with the AI engineers, and wire them into the pipeline so a failing evaluation blocks release.
• Run what you ship. Tracing, metrics, alerting, cost and drift monitoring.
• Review and pair. Design and code review, and pairing with early-career AI engineers who will often be ahead of you on AI craft; you are expected to be pulled forward. Templates and gateway access so AI tools built by the business run on the supported path.

The role

We have been mandated exclusively by one of our client, an important Asset Manager, for a Senior Software Engineer Role – AI Lead. In a dynamic team, they are building the future of the company ain terms of Technology and what they build sits on their data platform and increasingly uses AI. In terms of daily business, you will bring what keeps software alive after the demo: design, tests, observability, security, and the anti-patterns you have already paid for. The person we are looking for is an Hands-on senior engineer who can work with the platform engineer, the data platform, and the portfolio managers and analysts who write their own code.

In terms of description of what will be the position, here are some details of what you will need to do:

• Ship. Work with portfolio managers, analysts and operations to take use cases to production. Small scope, production quality.
• Set the code standard. Strict type checking and linting as CI gates, contract tests at service boundaries, pinned dependencies with automated updates, mandatory review, decision records for choices that are hard to reverse.
• Own delivery for what the team ships. Infrastructure as code, CI/CD, container builds, environments, releases and rollback. The platform engineer owns accounts, clusters and networking; you own everything from the container image up.
• Build security into the pipeline. Secrets management, least-privilege IAM roles for the services you ship, dependency and code scanning, SBOM and signed artefacts, policy as code. Four-eyes approval on every production change, enforced by the pipeline. Changes to pipelines, policies and IAM reviewed by the security team; repository and CI admin rights sit outside the engineering team; production access is break-glass only, logged and reviewed after use.
• Assemble the shared AI components, adopting before building: LLM gateway (model access, data residency, logging, cost, evaluation hooks); retrieval with access control enforced at query time; agent runtime with tool calling and human-in-the-loop validation; prompt and configuration versioning.
• Make evaluation a release gate. Build and run evaluations with the AI engineers, and wire them into the pipeline so a failing evaluation blocks release.
• Run what you ship. Tracing, metrics, alerting, cost and drift monitoring.
• Review and pair. Design and code review, and pairing with early-career AI engineers who will often be ahead of you on AI craft; you are expected to be pulled forward. Templates and gateway access so AI tools built by the business run on the supported path.

What success looks like in year one

• A first product in daily use by business teams, the previous way of working retired, adoption measured.
• The LLM gateway is live and every AI call from software we build goes through it; residency and logging are verifiable by the CISO.
• Every service the team runs deploys through one audited pipeline: infrastructure as code, security scanning, evaluation gate, tested rollback.
• Changes from early-career engineers pass your review without rework.

Requirements

In terms of profile, the chose one wil have the following experience and qualities:

• 8+ years building and operating production software others depend on, most of the last two spent writing production code. Python to a high standard: typing, packaging, testing, profiling. One statically typed language (Go, Kotlin, Java, TypeScript) used for real services, not scripts. API and data-contract design; SQL.
• You have written the infrastructure as code (Terraform, CDK or equivalent), the pipelines and the Dockerfiles yourself, on a public cloud, rather than requested them from another team. You have deployed and operated your own services on Kubernetes.
• You have implemented secrets management, least-privilege IAM and dependency scanning in a pipeline you owned, and threat-modelled a system you built, including prompt-injection and data-exfiltration paths.
• You have shipped and operated a production LLM system with retrieval or tool calling: you wrote the code, measured its quality, and fixed a failure mode users hit.
• You use AI coding agents daily, can say precisely what they are bad at, and have set rules for how a team uses them.
• You have defined alerts and SLOs for a service you built, and been on call for it.
• You have mentored an early-career engineer, and written a standard other engineers adopted and CI enforced.

Strongly preferred

• AWS in production: IAM, VPC networking, EKS or ECS, Bedrock.

Nice to have

• Led a small engineering team while still writing production code.
• Financial services or another regulated environment.
• Supply-chain security in practice: SBOM, artefact signing, policy as code (OPA, Checkov).
• GitOps (Argo CD, Flux); MCP and agent frameworks.

Benefits

Nice opportunity