Senior Software Engineer (Backend → AI Applications) - Remote, Worldwide

Are you the one?

Own and ship backend systems that power AI features. You’ll take ambiguous product goals, design reliable services, integrate LLM/ML components, and deliver production outcomes with minimal supervision. You already have a strong backend track record and have demonstrably upskilled in LLMs/ML (projects, repos, papers, or shipped features), and you’re excited to become a go-to engineer for AI applications.

What you’ll do

• Design & build services: Architect and implement production APIs, workers, and data flows for AI-powered features (auth, rate limiting, retries, idempotency, ops runbooks).

• LLM/ML integration: Ship features using model APIs and open-source models; implement tool/function calling, structured outputs, streaming, and robust error handling.

• Retrieval & data: Stand up RAG pipelines (chunking, embeddings, vector stores), quality checks, and freshness/consistency workflows.

• Evals, reliability & safety: Add offline/online evals, guardrails, red-teaming hooks, and observability (traces, token/latency metrics, prompt/version tracking).

• Performance & cost: Profile latency, memory, and token usage; add caching, batching, and fallbacks; keep cost/SLA within targets.

• MLOps basics: Containerize/model-serve where needed (e.g., vLLM/Triton), manage model/config/versioning, automate CI/CD.

• Ownership: Lead projects end-to-end with crisp written docs, RFCs, and clear stakeholder updates; mentor peers and raise the engineering bar.

Qualifications

• Proven senior-level backend experience (typically 4–8 years shipping services at scale).

• Strong in one of: Python or TypeScript/Node; plus familiarity with a systems language (Go/Java/C++) is a bonus.

• Solid distributed systems fundamentals: concurrency, queues/streams, retries, backoff, consistency trade-offs.

• Datastores: SQL (Postgres/MySQL), caching (Redis), and comfort with vector databases (e.g., pgvector/FAISS/Pinecone) or willingness to learn fast.

• Cloud & DevEx: Docker/Kubernetes, AWS/GCP/Azure, CI/CD, IaC (Terraform), and production monitoring (OpenTelemetry/Prometheus/Grafana).

• Demonstrated LLM/ML learning: shipped features or serious side projects using OpenAI/Anthropic/Mistral or open models; familiarity with LangChain/LangGraph/LlamaIndex (or equivalent), embeddings, and RAG.

• Product instincts and autonomy: you turn fuzzy goals into small, testable increments and communicate trade-offs clearly.

Nice to have

• Experience with eval frameworks, prompt/program design, JSON schema/structured outputs, and multi-agent orchestration.

• Model customization familiarity (LoRA/fine-tunes/distillation) and when not to train; basics of dataset curation.

• Model serving & performance: vLLM, tensor/kv-cache tuning, batching, quantization.

• Security & compliance: PII handling, GDPR, data governance.

• Interest in blockchain systems/Web3 (some clients are institutions and blockchain companies).

• OSS contributions, technical writing, or public demos of AI features.

Success metrics

• Delivery: Features shipped to prod on time with clear rollouts and rollback plans.

• Reliability: P95 latency and error budgets met; eval pass-rates and regression alerts in place.

• Quality & cost: Meaningful lift on target KPIs (e.g., resolution rate), with cost per request within budget.

• Ownership: High-quality docs/runbooks; peers adopt your patterns.

First 90 days (example)

• 30 days: Ship a scoped AI feature (e.g., RAG-backed endpoint) with logging, tracing, and basic evals.

• 60 days: Harden reliability (guardrails, fallbacks), add caching/batching, and cut cost/latency by ~20–30%.

• 90 days: Lead a second initiative (e.g., multi-tool agent or fine-tune evaluation), publish internal RFC + runbook.

What We Offer/ Benefits

• Flexible Work Options: Remote-first culture with the opportunity to work from anywhere.

• Global and Diverse Workforce: You'll work with people from various backgrounds and cultures.

• Learning and Development: You'll work on innovative, challenging projects and have access to experts and mentors to enhance your skills.

• Career Growth: Access to training, mentorship, and opportunities to contribute to open-source initiatives.

• Global Events and Conferences: Opportunities to attend the industry events.

• Collaborative and Innovative Culture: We foster teamwork and encourage new ideas.

Our Commitment to Diversity

At Nethermind, we celebrate diversity and are committed to creating an inclusive environment for all team members. We believe a variety of perspectives drives innovation and leads to better solutions for the blockchain community.

Ready to Join Us?

If you're passionate about blockchain and eager to make an impact, we’d love to hear from you. Click Apply for this Job to start your journey with Nethermind.