About the company

​

We’re an AI startup that ships AI-native products to highly regulated sectors—think banks, hospitals, and public-sector agencies where an audit trail matters as much as a killer UX. Our stack blends open-source LLMs, vendor APIs, and our own secret sauce to give enterprises “wow” moments. We run lean, bias to action, and celebrate prototypes that make it into production.

​

About the role

​

You’ll be the Founding Engineer on a small, senior team. One week you might wire a Bedrock-hosted RAG service; the next you’re pair-designing an adaptive UI for clinicians, or untangling IAM policies so a finance client’s CISO can sleep at night. You’ll own multiple tracks end-to-end—ideation, proof-of-concept, SOC-2-friendly hardening, and scaled rollout—while steering product, design, and data loops so they talk to each other in real time. Ambiguity is the norm; autonomy is the reward.

​

What we’re looking for

​

• Technical depth: CS/EE/Stats degree (or equivalent “dropped-out-to-ship” story) and fluency in Python, TypeScript, basic MLOps, and infra-as-code​

​

• Model deployment chops: you’ve pushed hosted or fine-tuned models to AWS (SageMaker, Bedrock, ECS) or Azure (ML, OpenAI) and know the trade-offs​

​

• Backend instincts: comfortable with container orchestration, event-driven services, blue-green or canary deploys, and observability hygiene​

​

• Regulated-env awareness: familiarity with SOC 2, HIPAA, PCI, or similar and how they influence data architecture
  and UX

 

What you’ve done

​

• Shipped at least one AI-powered product from zero to thousands of concurrent users— either in an internship or as a side project or even an academic deliverable​

​

• Built scalable services (K8s, Fargate, or serverless) that stayed upright under load, with latency budgets you
  actually met​

​

• Integrated LLMs or vector search into customer-facing workflows, then watched the metrics to iterate

 

Technical stack

​

• Cloud & ML AWS (Bedrock, SageMaker, Lambda, Fargate), Azure (ML, OpenAI, Functions), occasional self-hosted GPUs on Kubernetes

​

• Languages Python (⁠FastAPI, LangChain), TypeScript (Node/React), a dash of Go/Rust for latency-sensitive bits

​

• Data & Storage Postgres, DynamoDB, Redis, S3, Pinecone/pgvector for embeddings, Kafka/EventBridge for async flows

​​

Expectations

​

• Remote with in-person hybrid time

​

 

If you’re itching to turn ambiguous “what-ifs” into regulated-ready AI products,

please contact cpiovesan@inq.consulting.