AI Engineer (Agentic AI & Regulatory Reporting)

AI Engineer (Agentic AI & Regulatory Reporting)

Apexon is an AI-first technology services firm that engineers intelligent enterprises. Our interconnected disciplines of Domain & Strategy, Cognitive Architecture, and Harness Engineering guide every engagement, helping clients deliver meaningful business outcomes while applying collective intelligence to confidently scale from early AI strategies to enterprise-grade production systems. For over three decades, we have helped organizations by leveraging our deep and practical expertise in Artificial Intelligence, Data & Analytics, Digital Engineering, and Experience, combined with our industry knowledge across Banking & Financial Services, Healthcare and Life Sciences to create lasting business value. Apexon is backed by Goldman Sachs Asset Management and Everstone Capital.

We are looking for handson AI Engineers to build enterprisegrade GenAI and Agentic AI solutions for complex financialservices workflows. The role will focus on AIassisted exception classification, rootcause analysis, impact assessment, workflow orchestration and remediation across largescale transaction and dataprocessing environments.

The engineer will work with structured and unstructured enterprise data, historical issues, workflow/ticketing systems, business rules and operational context. Strong Python, SQL, API and softwareengineering fundamentals are essential, along with practical experience building LLM and agentic AI applications.

Key Responsibilities

Build AI workflows for exception classification, triage, rootcause analysis, impact assessment and remediation recommendations.

Develop agentic workflows involving orchestration, tool/function calling, state management and humanintheloop validation.

Correlate new exceptions with historical issues, known root causes, business rules and transaction/data attributes.

Build Python and SQLbased services to query, transform and analyze structured enterprise datasets.

Integrate AI applications with enterprise APIs, databases, workflow/ticketing platforms and internal data sources.

Develop reusable tools and services that AI agents can invoke for retrieval, investigation, analysis and workflow execution.

Apply RAG/context retrieval where regulatory documents, historical knowledge or issue repositories need to be searched.

Implement confidence scoring, validation, guardrails and traceability for AIgenerated outcomes.

Build backend services and APIs using Python/FastAPI.

Implement testing, logging, evaluation, observability, exception handling and production engineering practices.

Collaborate with onshore AI engineers, architects, business analysts, data engineers and application teams.

Skills -

Python

SQL & Data Analysis

GenAI / LLM Application Development

Agentic AI / Workflow Orchestration

APIs / FastAPI

Tool / Function Calling

Structured Data Processing

Problem Solving / RootCause Analysis

Git / Software Engineering

RAG / Retrieval

LLM Evaluation / Guardrails

Cloud. Agentic AI Experience

Handson experience with at least one agent/workflow orchestration framework such as LangGraph, Google ADK, CrewAI, AutoGen, Semantic Kernel, or an equivalent custom framework.

Candidates should understand how to implement workflows such as: Input Classification Tool/Data Retrieval Investigation Reasoning Validation Action.

Preferred Domain Experience

Financialservices experience is preferred but not mandatory.

Exposure to Regulatory Reporting,

Capital Markets, trade lifecycle, posttrade processing, transaction reporting, reconciliations, exception management, risk or controls will be advantageous.

Familiarity with regulations such as EMIR, MiFID II or SFTR is a plus.

Nice to Have

Model Context Protocol (MCP) and reusable agent tool interfaces.

Knowledge Graph / Graph RAG or datalineage concepts.

Vector databases, hybrid search or reranking.

AI observability and evaluation frameworks.

SpecDriven Development (SDD): ability to translate business/technical requirements into clear specifications, tasks and acceptance criteria before implementation.

Experience using Claude Code or equivalent AI coding assistants within disciplined softwareengineering practices.

Basic React or frontend integration experience

We are looking for engineers who can build the complete workflow around an AI modelnot simply prompts or basic chatbots. The candidate should be comfortable working across data, tools, agents, deterministic logic, validation, APIs and production engineering.

A representative problem could involve receiving a new reporting exception across a large transaction population, determining whether it maps to a known issue, identifying potentially impacted transactions, establishing supporting evidence, and routing unresolved cases for further investigation.

Qualifications

58 years of professional experience in software engineering, AI/ML engineering, data engineering, GenAI or related .