Software Engineer - Model Performance
Join Atmantara's core engineering team to drive performance across AI model deployments in fintech use cases.
Atmantara builds foundational AI infrastructure for financial institutions. From real-time fraud detection to multi-model portfolio optimization, we empower financial teams with scalable, secure, and compliant AI capabilities, without the engineering overhead. Our platform supports cloud-native and hybrid deployments with built-in observability, role-based access, and regulatory-grade compliance (SOC 2, GDPR, PSD2, MiFID II).
We're a high-velocity team based in Berlin and remote, working at the intersection of AI, infrastructure, and fintech.
You will work on the performance, scalability, and latency of AI workloads powering real-time decisions in financial systems. From optimizing inference pipelines to tooling for model evaluation and deployment, you'll build systems that deliver fast, reliable outcomes for our customers.
You'll partner with infra and product teams to ensure models perform under real-world fintech constraints: latency, concurrency, throughput, and compliance.
- Optimize inference serving infrastructure for LLMs and classification models
- Build observability tools to monitor model latency, throughput, and performance
- Design benchmarks and test frameworks to validate model performance
- Support deployment of models in production environments across hybrid infra
- Work with research engineers to close the loop between model design and deployment
- Strong background in backend or systems programming (Go, Rust, Python, or similar)
- Experience with ML inference frameworks (TorchServe, TensorRT, Triton, etc.)
- Comfortable working on latency-sensitive, production-grade systems
- Experience with containers, orchestration (Kubernetes, Docker, etc.)
- Familiarity with observability stacks (Prometheus, Grafana, OpenTelemetry)
- Exposure to high-throughput financial systems or trading infrastructure
- Built custom model serving or scaling infra at a previous company
- Experience with asynchronous programming, queues, and batching systems
- Contributions to open-source model deployment or performance tooling
- Competitive salary and early-stage equity
- Flexible remote-first setup with a central hub in Berlin
- Annual learning & development stipend
- Parental leave and dedicated wellness budget
- Covered public transportation or mobility allowance
- Company laptop and full remote workstation setup
- Friday after-work dinners and drinks
- Annual team offsites and international retreats
- Regular guest talks from AI and fintech industry leaders
- Dedicated focus days, no internal meetings, deep work encouraged
We believe AI infrastructure should accelerate financial intelligence. Atmantara empowers lean, focused teams to work with autonomy and deep collaboration. We measure success in real outcomes: helping financial institutions deploy smarter models faster, without bottlenecks or complexity.
If you're excited to help build the platform powering the next generation of AI in finance, we'd love to talk.
By applying, you agree to our Terms and Privacy Policy