LLM Application Engineer

• Own production LLM pipelines end to end, including ingestion, multimodal model calls, structured records, storage, confidence flags, retries, and idempotent rescans
• Design prompt and schema strategies, including schema-aligned or constrained outputs, for consistent product-ready results
• Build classification and filtering layers, including taxonomy mapping, demographic or audience filters, deduplication, and cleanup logic
• Define and run evaluation harnesses with golden sets, regression suites, and online metrics
• Report against quality targets for completeness, duplicates, incorrect inclusions, image presence, and related product SLAs
• Optimize token usage, model tiering, caching, and batching to control cost and latency
• Harden long-running asynchronous jobs with timeouts, partial recovery, memory limits, and safe production deployments
• Partner with Flutter/mobile and QA teams on field contracts, review queues, and incident debugging
• Document architecture and runbooks for shared ownership
• Stay current on Gemini and peer LLM APIs; recommend model swaps, fallbacks, and schema improvements

• Shipped LLM applications in production (not demos only): prompts, structured outputs, retries, observability, and real failure handling
• Strong hands-on experience with Google Gemini, including multimodal text, image, and document-style workflows and structured extraction
• Practical experience with at least one other major LLM stack, such as OpenAI or Anthropic
• Structured extraction from HTML, PDFs, images, and mixed email-like content
• Classification and taxonomy systems on top of LLM outputs
• Evaluation discipline: offline evals, regression suites, and production quality metrics tied to clear acceptance criteria
• Cost and latency awareness: token budgeting, cheaper tiers, caching, and batching
• Python backend experience on serverless cloud, such as Cloud Functions, and document stores, such as Firestore, or similar GCP patterns
• English at C1 or above
• US hours overlap through roughly 5 PM EST when live coordination is needed
• Ownership habits: honest estimates, early blockers, and finished releases
• Nice to have: schema-aligned LLM frameworks, Google Document AI or other OCR/document intelligence, Gmail API/OAuth compliance familiarity, computer vision, Firebase/GCP operations, production evaluation corpora, and agency or multi-client studio experience

• Fully remote and async-friendly, with required overlap through ~5 PM EST when client or release coordination needs it
• Monthly retainer structure with recurring AI pipeline work for engineers who keep production quality and cost honest
• Real autonomy over how you structure prompts, schemas, evals, and deploys
• Direct line to PMs, mobile engineers, and decision-makers
• Tooling budget for the LLM and cloud tools you need to move fast
• A peer network of product-minded builders across overlapping client projects