Lead Autonomous Agent & RAG Engineer
Design hybrid semantic search (BM25 + dense vectors), cross-encoder rerankers, and evaluation harnesses for production agentic workflows.
Base Compensation
$105k – $140k USD
Transparent, annualized global tier.
Work Arrangement
100% Remote
Async-first, any timezone worldwide.
Squad Density
100% Senior Eng
Zero junior overhead or hand-offs.
Hiring SLA
48-Hour Decision
Fast offer post-technical sprint.
Executive Overview & Mission
At GrowBig IT, retrieval is not a demo feature. It is the core of enterprise agent systems that must cite every answer, survive hostile documents, and hold latency budgets under load.
As Lead Autonomous Agent & RAG Engineer, you will own the retrieval stack end to end: chunking strategy, hybrid indexing, reranking, evaluation harnesses, and the feedback loops that keep quality from drifting in production.
Core Tech Stack & Autonomous Infrastructure
Technologies actively deployed in our current production client contracts:
- Python 3.12+ (Asyncio)
- LangGraph & LangChain Core
- Pinecone & Qdrant
- Cohere Rerank v3
- BM25 Hybrid Search
- OpenAI Structured Outputs
- Claude 3.5 Sonnet / Haiku
- Ragas & Promptfoo
- Redis & Celery
- Langfuse & OpenTelemetry
What You Will Architect & Ship
Hybrid Retrieval Pipelines: Combine dense embeddings with BM25 sparse signals and cross-encoder reranking for high-precision recall.
Chunking & Metadata Strategy: Design document pipelines that preserve structure, provenance, and citation integrity.
Evaluation Harnesses: Build golden datasets and automated scoring that catch regressions before clients do.
Agentic Query Planning: Route complex questions across tools, indexes, and multi-step graph traversals.
Cost & Latency Compression: Instrument retrieval traces and cut token spend without losing answer quality.
What We Are Looking For
5+ years of production software engineering in Python or high-scale TypeScript.
2+ years building retrieval or search systems with measurable quality ownership.
Fluency with vector databases, embedding models, and hybrid search tradeoffs.
Strong distributed fundamentals: async I/O, queues, caching, and containerized deploys.
Demonstrated shipped systems: public repos, papers, or deep architectural war stories.
Compensation & Sovereign Perks
Top-Tier Global USD Pay
$105,000 – $140,000 USD paid seamlessly through Wise, Deel, or direct global wire transfer.
$3,000 AI & Compute Stipend
Annual budget for OpenAI/Anthropic API research credits, GPU instances, and technical books/certifications.
Top-Spec Hardware Gear
Factory-fresh Apple MacBook Pro (M3 Max / M4 Pro, 64GB Unified RAM) + $2,000 home office workstation setup.
Annual Squad Retreats
28 days paid leave + premium health coverage + fully sponsored annual gatherings in destinations like Bali, Lisbon, and Dubai.
4-Step Zero-Fluff Hiring Process
No automated screening bots, no leetcode puzzles that bear no resemblance to real engineering.
- 1
Async Architecture & GitHub Review
We review your repository, PR history, or technical writing within 24 hours of submission.
- 2
Paid 2-Hour Practical Sprint ($200 USD Stipend)
A take-home architectural problem mirroring actual client edge cases. Compensated for your time.
- 3
45-Minute Deep-Dive with CTO Rafiqul Karim
Direct peer-to-peer code and systems review. Zero HR corporate fluff, strictly technical alignment.
- 4
Written Offer Within 48 Hours
Clear compensation, equity incentives, contract parameters, and rapid onboarding timeline.
Submit Direct Application
Direct review by CTO leadership within 24 hours. No recruiters, no keyword filters.
Active Squad Expansion
Other High-Velocity Openings
Principal Next.js & Full-Stack Architect
Lead high-concurrency builds in Next.js, TypeScript, and Postgres. Own architecture, edge caching, and design-system performance for flagship products.
Lead Autonomous Agent & RAG Engineer
Design hybrid semantic search (BM25 + dense vectors), cross-encoder rerankers, and evaluation harnesses for production agentic workflows.
Senior DevOps & Kubernetes SRE
Own 24/7 reliability for multi-tenant Docker, Traefik, Postgres, and Redis stacks. Automate zero-downtime releases and SOC 2 compliance infrastructure.