Ai Engineer - Document Intelligence

Hirexe; Remote; ZAR 60 000 - 80 000; Full time

Job summary

A technology company is seeking an AI Engineer to develop and enhance its AI layer, focusing on contract document intelligence. This role involves optimizing RAG implementations for document analysis and improving the parsing accuracy of legal documents. Candidates should have experience in LLM integration and document processing. The company offers a flexible work structure and is positioned in a massive market needing innovation. Join now to impact the development of AI-driven commercial real estate solutions.

Benefits

Flexible working hours

Impactful work in a growing market

Work with cutting-edge technology

Qualifications

• Hands-on experience designing and optimizing RAG systems, particularly for document analysis or Q&A.
• Production experience with OpenAI, Anthropic, or similar LLM APIs.
• Experience with parsing, extracting, and structuring data from unstructured documents.

Responsibilities

• Optimize and refine RAG implementation for contract analysis and recommendations.
• Improve document parsing accuracy across Word and PDF lease documents.
• Design strategies for handling complex, interdependent contract clauses.

Skills

RAG Pipelines

LLM Integration

Document Processing

Vector Databases

Prompt Engineering

Tools

AWS Lambda

FastAPI

OpenSearch

Python

React

Job description

Location :

Remote – Must be willing to work EST hours

About MarkIt

MarkIt is transforming how commercial real estate deals get done.

We're building AI-powered tools to solve one of the most painful, manual processes in CRE : lease negotiations.

Every deal involves principals, brokers, attorneys, and months of back-and-forth on documents that haven't evolved in decades.

We're turning a day process into hours and days while dramatically reducing costs.

Commercial real estate is the largest asset class in the world ($20+ trillion) and shockingly underserved by technology.

Our founder has been a top broker in the industry for over a decade and knows this pain intimately.

What We're Building

An intelligent platform that transforms how commercial real estate professionals negotiate lease agreements—combining collaborative workflows with AI-powered document analysis to accelerate deals and reduce costs.

The Platform

Purpose-built transaction management involving counter-party collaboration to improve upon

Workflow automation to streamline back-and-forth negotiations between principals, brokers, and attorneys

Enhanced features and organization compared to current workflows for all stakeholders negotiating transactions

AI-Powered Intelligence (Your Focus)

Proprietary RAG models that extract and parse complex lease agreements, understanding relationships between dozens of interdependent clauses

Context-aware recommendations for drafting and negotiating terms based on market standards and historical data

Intelligent drafting of business and legal language according to templates and / or market data

Where We Are

We're at an exciting 0.5?1 stage :

Product prototype :

Built and functional with users in pilot testing

Tech stack :

React / Next.js frontend, .NET backend, Python AI services on AWS

Timeline :

Entering UAT in the next two months, targeting MVP by Q2

Stage :

Pre-launch, pre-revenue, with strong customer interest

The Role

We're looking for an AI Engineer to own and evolve our AI layer.

You’ll work directly with our technical team and founder to push the boundaries of what’s possible with contract document intelligence.

Primary Responsibilities

• RAG & Document Processing (Core Focus)
• Optimize and refine our RAG implementation for contract analysis and recommendations
• Improve document parsing accuracy across Word and PDF lease documents
• Enhance semantic search and retrieval across legal clause libraries
• Design and implement strategies for handling complex, interdependent contract clauses
• LLM Optimization & Prompt Engineering
• Refine prompts and model configurations for legal language rewriting
• Develop strategic trade-offs between model performance and computational costs
• Experiment with different LLM providers and models (we currently use AWS Bedrock)
• Implement evaluation frameworks to measure recommendation quality
• Agentic AI & Innovation
• Discover and implement innovative agentic AI workflows for negotiation assistance
• Explore multi-step reasoning approaches for contract analysis
• Build systems that can explain recommendations and cite supporting evidence
• Push the boundaries of what’s possible with contract document AI
• Work closely with our technical team on API integrations and data pipelines
• Provide technical guidance on AI architecture and model selection
• Own the AI roadmap and help prioritize feature development
• Communicate complex technical concepts to non-technical stakeholders

AI Infrastructure

Document Processing :

Python on AWS Lambda, AWS Bedrock, DynamoDB, S3

Document Analysis :

Python FastAPI on AWS ECS, OpenSearch, AWS Bedrock

Models :

Currently using Claude (Anthropic) and other models via Bedrock

Vector Search :

OpenSearch for semantic retrieval and chat history

Supporting Stack (nice to have familiarity, not required)

Testing :

Playwright (e2e), Storybook

What We're Looking For

Required Experience

• RAG Pipelines : Hands-on experience designing and optimizing RAG systems, particularly for document analysis or Q&A
• LLM Integration : Production experience with OpenAI, Anthropic, or similar LLM APIs
• Document Processing : Experience with parsing, extracting, and structuring data from unstructured documents (PDFs, Word docs, contracts, legal documents)
• Vector Databases : Experience with vector search systems (OpenSearch, Pinecone, Weaviate, etc.)
• Prompt Engineering : Demonstrated ability to craft and optimize prompts for complex reasoning tasks

Highly Valued Experience

• Contract / Legal Document AI : Prior work on contract review, legal document analysis, or compliance systems
• AWS : Experience with Lambda, ECS, Bedrock, or other AWS AI services
• FastAPI : Building and deploying Python APIs
• Model Evaluation : Frameworks for measuring LLM output quality and accuracy
• Backend Development : .NET, Node.js, or similar for supporting additional development needs
• MLOps : Experience with model monitoring, versioning, and deployment pipelines
• LangChain / LlamaIndex : Familiarity with LLM application frameworks
• Fine-tuning : Experience with model fine-tuning or customization
• Real Estate / Legal Domain Knowledge : Understanding of contracts, leases, or commercial real estate

Nice to Have

• Backend Development : .NET, Node.js, or similar for supporting additional development needs
• MLOps : Experience with model monitoring, versioning, and deployment pipelines
• LangChain / LlamaIndex : Familiarity with LLM application frameworks
• Fine-tuning : Experience with model fine-tuning or customization
• Real Estate / Legal Domain Knowledge : Understanding of contracts, leases, or commercial real estate

What We Offer

Role Structure

• Start : Flexible part-time or full-time opportunity based on preference and availability

Why This Is Exciting

Ground floor opportunity : Shape the product, technical architecture, and culture from day one

Massive market : $20+ trillion asset class desperate for innovation

Real problem, real users : Founder knows the problem intimately; strong customer interest already

AI at the core : This isn't AI as a feature—it's the entire product.

Work with cutting-edge LLM technology on genuinely hard problems

Autonomy : Small enough that you own outcomes, not just tasks.

Your decisions matter immediately

Impact : The next 90 days will determine our MVP timeline and market advantage—you'll be central to that

About Our Founder

Jason White has been a top commercial real estate broker at one of the world's leading CRE firms for over a decade, giving him deep knowledge of the negotiation pain points we're solving.

He's represented more than half of the Fortune 10 as well as hundreds of local and regional companies throughout his career.

This is a founder who's lived the problem and is building the solution with the right team.

Seniority level

Associate

Employment type

Full-time

Job function

Engineering and Information Technology

Industries

Software Development

#J Ljbffr