AI Agents Engineer – Python, FastAPI, RAG & Tool Calling

MaxIT Consulting - Max Corporate Group; Los Altos (CA); On-site; USD 120,000 - 190,000; Full time

Job summary

MaxIT Consulting is seeking a Backend Engineer, Applied Agents for an on-site role in Los Altos, CA. You will design, build, and operate the agent layer and backend services powering AI workflows in drug development.

Collaborate with founders and customers to ensure reliable, verifiable outcomes. You will implement the agent runtime, tool-calling, sandboxed execution and scalable data pipelines using Python, FastAPI and Postgres, while focusing on reliability, observability and latency

Qualifications

• 2+ years shipping and operating production backend systems end to end.
• Strong fundamentals in API design, data modeling, databases and system reliability.
• Hands-on production experience with LLM systems: RAG, vector databases and agents.
• Has built, or deeply traced, an agent runtime or harness: tool calling, sandboxing, memory.
• Ships fast on incomplete specs and uses AI coding tools (Claude Code, Cursor) daily.

Responsibilities

• Build the agent runtime and harness, tool-calling infrastructure, sandboxed code execution and agent memory.
• Own reliability and observability of multi-step agent tasks.
• Design and build backend services and APIs on Python, FastAPI and Postgres.
• Design and operate vector databases, semantic search and RAG systems, tuning quality and latency.
• Build large-scale ingestion, cleaning and indexing pipelines for scientific and regulatory documents.
• Ground AI outputs in source data with traceable, verifiable citations.
• Work directly with founders, domain experts and early customers, owning outcomes.

Skills

API design

Data modeling

Databases

System reliability

LLM systems

RAG

Vector databases

Agents

Agent runtime

Memory

Tools

Python

FastAPI

Postgres

Job description

Los Altos, CA (San Francisco Bay Area) | On-site | Full-time

MaxIT Consulting is looking for a Backend Engineer, Applied Agents to join a venture-backed, early-stage AI company building an AI-native operating system for drug development programs, already used by leading biopharma companies.

The Opportunity

You will design, build and operate the agent layer and the backend systems that power it: agents that carry out multi-step tasks inside real workflows, call tools reliably, run code in sandboxed environments, keep context and deliver verifiable results.

Key Responsibilities

• Build the agent runtime and harness, tool-calling infrastructure, sandboxed code execution and agent memory.


• Own reliability and observability of multi-step agent tasks.


• Design and build backend services and APIs on Python, FastAPI and Postgres.


• Design and operate vector databases, semantic search and RAG systems, tuning quality and latency.


• Build large-scale ingestion, cleaning and indexing pipelines for scientific and regulatory documents.


• Ground AI outputs in source data with traceable, verifiable citations.


• Work directly with founders, domain experts and early customers, owning outcomes.

Required Qualifications

• 2+ years shipping and operating production backend systems end to end.


• Strong fundamentals in API design, data modeling, databases and system reliability.


• Hands-on production experience with LLM systems: RAG, vector databases and agents.


• Has built, or deeply traced, an agent runtime or harness: tool calling, sandboxing, memory.


• Ships fast on incomplete specs and uses AI coding tools (Claude Code, Cursor) daily.

Nice to Have

• Built an in-house agentic harness or internal tools to boost a team's agentic development.


• LangGraph or agent frameworks in production; AWS, Terraform and Kubernetes.


• Enterprise SaaS, healthcare or life sciences products; seed or early-stage startup experience.

Work Arrangement

On-site role in Los Altos, CA (San Francisco Bay Area); fully remote is not possible. Candidates must live in or relocate to the Bay Area and must already be authorized to work in the United States.