AI search is changing how brands are discovered and chosen. People now ask ChatGPT, Gemini, Claude and Google’s AI answers what they used to search for, and AI assistants increasingly shape the journey from discovery to purchase. Winning customers now means winning over the AI systems that influence their decisions.
Peec AI is an AI search platform that helps marketing teams understand and improve how their brands are discovered, represented, and recommended. We show them where they stand, what’s shaping it, and where to act. Since launching in January 2025, Peec AI has passed $20M ARR and 4,000 paying customers. We’re a Series A company backed by 20VC and Singular, with 100+ people across our Berlin HQ and our New York office.
Peec AI is building the data plane for AI visibility: millions of chats, scrapes, citations, sources, model responses, and customer-facing insights running through a fast-moving backend.
This role is for an engineer who wants to own the systems behind that. You will not just add endpoints or close tickets. You will make our scraping, chat-processing, source-ingestion, and analytics infrastructure more reliable, cheaper, faster, and easier for the rest of the team to reason about.
The best fit looks like someone who can debug a provider outage in the morning, design a better queueing model before lunch, ship a TypeScript service change in the afternoon, and write the postmortem or follow-up ticket clearly enough that the next engineer immediately knows what to do.
We are especially interested in people who have built or operated systems with real constraints: rate limits, flaky third-party APIs, queues, webhooks, retries, distributed locks, data migrations, observability gaps, customer pressure, and cost ceilings.
5+ years of backend, platform, infrastructure, or distributed systems engineering experience in a production environment.
Deep hands-on TypeScript experience. We run TypeScript services on Bun, and we care about people who can design clean, reliable backend systems rather than only write application code.
Strong understanding of queues, workers, retries, rate limits, idempotency, distributed locks, webhooks, and failure handling.
Experience operating systems under real production pressure: incidents, provider outages, customer escalations, noisy alerts, slow dependencies, and incomplete data.
Comfort with cloud-native infrastructure. Our stack includes GCP, Kubernetes, Cloud Tasks, Pub/Sub, RabbitMQ, Redis, Postgres, Firestore, ClickHouse, Prometheus, Grafana, and OpenTelemetry.
Ability to debug across layers: application code, logs, metrics, queues, provider APIs, databases, network behaviour, and deployment state.
Product-minded judgement. You can separate internal detail from customer-safe explanation, and you understand when a “small infra change” has customer, security, or cost implications.
High ownership and low ego. We want people who take responsibility for ambiguous systems, ask sharp questions, and make the team better through clear thinking and kindness.
A bias toward durable fixes. You do not stop at “restart the pod” if the real issue is retry amplification, missing backpressure, an unsafe parser assumption, or a provider limit we do not model.
Languages: TypeScript, Python
Runtime/frameworks: Bun, Elysia, Node-compatible services
Cloud/infrastructure: GCP, Kubernetes, Cloud Tasks, Pub/Sub, Firebase/Firestore, Postgres, ClickHouse, Redis, RabbitMQ
Observability: Prometheus, Grafana, OpenTelemetry, structured logging
Product/data domain: LLM chats, AI search, source scraping, citation extraction, provider integrations, customer-facing analytics
AI/model providers: OpenAI, Anthropic/Claude, Perplexity, Gemini, Google AI Mode / AI Overview, plus scraping/provider APIs around them
Exciting and challenging work with real impact and ownership at one of Europe’s fastest-growing Series A startups
Regular team events and off-sites
Aggressive equity compensation package
Paid Dinner & Uber home when working late
The most beautiful office space and work environment in Berlin