Lead AI Engineer - Conversational Intelligence & Memory Architecture

This role is US-based only. We do not offer visa sponsorship. No agencies or offshore applicants.

You must be legally authorized to work in the United States.

Role: AI Engineer (Founding / Lead)

Compensation: $75/hour

• IF REMOTE WORK: Screen activity tracking or recording is required during billed hours to ensure transparency and alignment.
• IF LOCAL TO MIAMI: Miami-based candidates must complete billable hours in-office.

WE ARE LOOKING FOR SOMEONE ABLE TO MOVE TO MIAMI OR CURRENTLY LIVING IN MIAMI AND WILLING TO BE IN THE OFFICE AT A MINIMUM OF 3 DAYS A WEEK.

About Capsule

Capsule is building an AI system that captures, organizes, and interacts with human memory, personality, and voice. Our AI interviewer (“Remi”) conducts guided reminiscence, while Capsule becomes a living, evolving model of a person—shaped by stories, preferences, relationships, and recorded voice.

This is not a toy chatbot. This is identity, memory, and legacy—and it requires taste, rigor, and care.

The Role

We’re looking for a senior AI builder who wants to step into a CTO-type role and architect Capsule’s intelligence layer end-to-end. You’ll own how memory is captured, structured, retrieved, and experienced—and how the system improves over time through evaluation, iteration, and production learning.

You’ll be hands-on: designing the architecture, writing code, shipping to production, and partnering directly with the founder on roadmap and product decisions.

What You’ll Do

• Architect the intelligence layer powering Remi and Capsule–includes memory ingestion, structuring, retrieval, and long-term context.
• Build the memory system (schemas/representations, embeddings + retrieval, summarization, entity/relationship tracking, user controls).
• Improve conversation quality: coherence, emotional attunement, safety, and consistency across long time horizons.
• Design and run evaluation: offline test sets, golden conversations, automated scoring, regression tests, and online quality monitoring.
• Own experimentation to production: prototype quickly, validate with users, deploy, monitor, iterate.
• Make infrastructure decisions: model/provider strategy, pipelines, observability, latency/cost tradeoffs, data privacy and retention.
• Partner with product/design to translate human needs (trust, grief, family dynamics, legacy) into robust technical systems.
• Lead technically as the team grows: set standards, mentor, and help define hiring priorities over time.

What We’re Looking For

• Shipped an LLM product to production with real users (quality + reliability mattered).
• Hands-on with RAG/retrieval, embeddings, tool/function calling, and orchestration.
• Built evaluation/monitoring (offline + online) and iterated based on data.
• Strong engineering fundamentals (APIs, data modeling, security/privacy basics, observability).

Bonus Points

• Former technical founder or early-stage startup lead (0 to 1, shipping under constraints).
• Experience with voice (STT, diarization, voice cloning/synthesis, audio pipelines).
• Background in NLP, cognitive science, psychology, or human-centered AI.
• Familiarity with AI-assisted development workflows (Cursor/Codex or similar) to move faster.