AI Engineer

About the role

About Patlytics:

Patlytics is the fastest-growing AI-native patent intelligence platform, transforming how IP is protected and monetized at scale. Our advanced LLMs and generative AI engine—custom-built for intellectual property—power every phase of the patent workflow from invention disclosure through litigation, delivering citation-backed insights with unprecedented accuracy and speed.

Fresh off our $40M Series B — led by SignalFire https://www.linkedin.com/company/signalfire/, with participation from N47 https://www.linkedin.com/company/n47capital/, Myriad Venture Partners https://www.linkedin.com/company/myriad-vp/, Liquid 2 Ventures https://www.linkedin.com/company/liquid-2-ventures/, BAM Corner Point https://www.linkedin.com/company/bam-corner-point/, Alumni Ventures https://www.linkedin.com/company/alumniventures/, Antiportfolio Ventures https://www.linkedin.com/company/antiportfolio-ventures/ and Relativity https://www.linkedin.com/company/relativityhq/- allows us to triple down on the product, accelerate growth globally, hire future mission-aligned colleagues, and to continue redefining what is possible for our customers. — we're scaling fast across product, engineering, and go-to-market. If you want to build at the inflection point, this is it.

With teams across the globe we're building a global company fueled by exceptional talent and bold ideas. Our rapid expansion is driven by people who bring entrepreneurial energy, intellectual curiosity, and a shared commitment to setting the global standard for AI-powered IP intelligence. We're creating a culture where diverse perspectives shape breakthrough innovation and every voice contributes to transforming how the world protects ideas.

We hire for trajectory, not just credentials. If you're 70% of this list and hungry to close the gap, apply.

As an AI Engineer, you’ll:

- Develop and deploy robust, scalable AI/ML algorithms for cutting-edge IP applications

- Design, implement, and iterate on multi-step LLM pipelines for patent claim analysis, infringement detection, and prior art search, including the model selection, prompt architecture, and structured output contracts that define product quality

- Build and maintain a rigorous evals framework: define success metrics per pipeline stage, curate golden datasets from real IP cases, and run continuous regression testing to catch degradation before users do

- Develop retrieval systems (vector search, BM25 hybrids, re-ranking) optimized for patent corpus characteristics, long documents, technical claim language, biological sequence identifiers

- Architect and ship agentic workflows within our Agent layer, coordinating tool use, memory, and multi-turn reasoning over complex IP research tasks

- Develop data classification techniques and post-training on the latest LLMs

- Collaborate with patent attorneys and domain experts to encode expert judgment into prompts, rubrics, and fine-tuning datasets, then measure whether it worked Own cost, latency, and quality tradeoffs for production inference: prompt caching strategies, context window management, batching, and model migration planning on AWS, GCP and other LLM providers.