SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
DevRev is building Computer, an AI teammate that unifies data sources, tools, and workflows into a single AI-ready platform for enterprises. The company is ushering in the SaaS 2.0 era where enterprise software is intelligent and agentic, not just connected.
As a Software Engineer in AI Performance, you'll join a structured training track to develop expertise in AI benchmarking, profiling, and performance tuning. Your core responsibilities include building and scaling benchmarking infrastructure for evaluating AI systems in enterprise settings, designing agent evaluation pipelines that measure reasoning, accuracy, alignment, and user outcomes, and collaborating with AI and infrastructure teams to instrument model-serving systems (vLLM, Triton, Ray, Kubernetes).
You'll develop data and telemetry pipelines that capture model performance, latency, and quality metrics. You'll contribute to DevRev's open benchmarking frameworks, helping define how intelligent SaaS software is measured globally. Working cross-functionally, you'll ensure benchmarks reflect real-world customer value and system impact.
The role requires 3–10+ years as a Software Engineer in backend, systems, or infrastructure roles. You should have strong Python or similar backend language skills (Go, Java, C#, Node.js), hands-on experience building APIs and cloud-native architectures (AWS, GCP, Azure; Kubernetes preferred), and familiarity with metrics, profiling, or observability tools (Grafana, Prometheus, ELK). Prior AI experience is not required—strong engineering fundamentals, curiosity, and commitment to learning AI benchmarking and LLM evaluation are essential. Bonus skills include vector search, model tuning (LoRA, DPO), or open-source contributions.
DevRev is backed by Khosla Ventures and Mayfield with $150M+ raised and is trusted by global companies across industries.
