Senior Backend Engineer

About us:

CNTXT AI is a UAE-based Data and AI company that helps organizations become AI-ready by delivering sovereign, end-to-end AI solutions built from the region, for the region, and designed to scale globally.

We provide secure data services, custom AI solutions and industry-specific AI applications that remove the friction from AI adoption — helping enterprises and governments cut costs, accelerate deployment, and maintain complete control over their data.

With our Arabic-first AI innovation alongside sector-tailored solutions, CNTXT AI ensures organizations unlock AI's full potential — on their terms, with full sovereignty.

The Mission

We are building an AI-native Accounting Operating System. While traditional platforms are built for human data entry, our OS is designed for Agentic-speed workloads. Our users don't just input financial data but ask "What is my cash balance in 7 days if my top 3 clients pay late?" To answer this, our backend must simulate thousands of scenarios and process tens of thousands of transactions per second (TPS) while maintaining the mathematical truth of an immutable General Ledger. Technical Skills (The Toolbox) Core Backend & Language Java: Expert knowledge of the JVM, specifically focusing on Virtual Threads to handle massive I/O concurrency. High-Throughput Persistence: Expert-level PostgreSQL. Proficiency in lock-less concurrency, MVCC, and partitioning. Familiarity with high-TPS distributions is highly preferred. Messaging & Streaming: Kafka or Redpanda for append-only eventsourcing and building real-time Change Data Capture (CDC) pipelines.

The "Accounting OS" Specialized Skills

Ledger Logic: Deep understanding of Double-Entry constraints (Atomic Debit/Credit pairings). Temporal Modeling: Experience building Bi-temporal or Tri-temporal data models (tracking Transaction Date vs. Posting Date vs. System Entry Date). Precision Math: Strict adherence to BigDecimal usage and multi-currency rounding standardizations (ISO 4217).

Immutability: Designing append-only schemas where corrections are handled via reversal entries rather than UPDATE or DELETE statements.

AI & Data Engineering

Agentic Orchestration: Experience building high-speed "Data-toInference" pipes, enabling LLMs to perform sub-second lookups across millions of ledger rows. Vector Readiness: Understanding of how to structure financial metadata for Vector search/RAG (Retrieval-Augmented Generation).

Key Responsibilities

1. High-Performance Transactional Engine Scale for AI-Inference: Design a ledger engine capable of 1,000+ TPS, ensuring that ad-hoc AI prompts can "fan out" into thousands of sub-calculations and "What-If" simulations in milliseconds. Shadow Postings: Engineer "Shadow Ledgers" such as temporary, highspeed sandboxes where AI agents can simulate financial forecasts without impacting the production General Ledger. Consistency Guardrails: Implement database-level constraints and application-level checks to prevent unbalanced journals or negative cash scenarios. 2. Temporal & Audit Integrity Period Control: Build robust "Hard Close" and "Soft Close" mechanisms to lock fiscal calendars while allowing AI models to forecast across periods. Immutable Audit Trail: Design a system where every single byte of financial data is traceable, non-destructive, and auditor-ready (SOC2/IFRS compliance).

3. Production Excellence

CQRS Implementation: Separate "Write" paths (Ledger Postings) from "Read" paths (AI Analytics) to ensure complex prompts never block realtime financial operations.

Test-Driven Development (TDD) for Finance: Maintain near-100% test coverage for core financial logic using property-based testing to catch mathematical edge cases.

Predictive Latency: Optimize database indices and caching layers (Redis/Dragonfly) to ensure "Agent-speed" responses to natural language prompts.

Required Qualifications

Experience: 5+ years in backend engineering, with at least 2 years in Fintech, Banking, High-Frequency Trading, or similar.

Systems Thinking: The ability to explain how to prevent deadlocks in a high-concurrency SQL environment.

Education: BS/MS in Computer Science, or equivalent experience (strong mathematical background is a plus).