Core BackEnd  AI  Integration  Engineer (Nest.js • Multi‑Model Orchestration • Social‑Platform APIs)

Full Description

Role summary

Provide the rock‑solid, observable back‑end that orchestrates LLM, vision, and voice calls, enforces API quotas, and exposes blazing‑fast REST/GraphQL endpoints. You’ll own Nest.js micro‑services, automatic retries, and database design—keeping latency and cost predictable at scale.

Core responsibilities

What you’ll do

• Service design: Model a Nest.js monorepo (gateway, workers, contracts), layer CQRS, and auto‑generate OpenAPI/GraphQL SDL.

• AI orchestration: Build a vendor‑agnostic execution graph, stream results, and trace cost/latency via OpenTelemetry.

• RAG & embeddings: Index corpora in pgvector on Supabase; combine BM25 & HNSW and stream retrieval chunks via SSE.

• Social‑media adapters: Implement OAuth 1.0a/2.0 (PKCE) for Twitter/X, verify webhooks, and back‑pressure events through Redis Streams.

• State & storage: Use MongoDB for schemaless chat logs, Postgres for relational entities, Redis for queues & rate limiting.

• Dev‑excellence: Cover critical paths with Jest & Pact, ship Docker images through GitHub Actions, and smoke‑test staging on Render/ECS.

Must‑have skills

• 3 + yrs Node/Nest.js in production (decorators, interceptors, CQRS).

• Deep rate‑limiting, retry, and idempotency‑key expertise.

• Production integrations with at least two LLM providers (streaming & tool‑calling).

• Comfortable with Redis Streams, MongoDB replicas, and Postgres/Supabase.

• Strong REST/GraphQL craftsmanship (auth, pagination, streaming).

Bonus skills

• LangChain or LlamaIndex in production (custom retrievers, memory).

• Deploy pipelines on AWS ECS/Fargate or Render.

• Experience with Solana Agent Kit or Eliza OS contract calls.

• Familiarity with OpenTelemetry/Tempo and cost‑tracking for multi‑vendor AI usage.

Job Type: Full-time

Work Location: Remote