Senior AI Engineer

FinacPlus; Bengaluru; On-site; INR 3,500,000 - 5,500,000; Full time

Job summary

FinacPlus in Bengaluru is seeking an experienced AI/ML engineer to design, build, and operate AI-powered systems for mortgage and structured credit documents. You will work on classification, extraction, splitting, and compliance tagging using LLMs and OCR, and help embed AI tooling across a 50-engineer org.

This dual-track role blends product-facing AI features and platform engineering for AI adoption, with focus on MLOps, observability on GCP, and driving analytics-led improvements.

Qualifications

• Experience designing end-to-end AI features from data ingestion to deployed API.
• Strong knowledge of LLMs, OCR, and RAG workflows.
• Experience with MLOps, model versioning, A/B testing, and observability.
• Familiarity with cloud platforms (GCP) and BI tooling for analytics.

Responsibilities

• Design and maintain intelligent document pipelines for mortgage artefacts — loan packages, appraisals, title reports, closing disclosures — covering classification, splitting, field extraction, and audit tagging using LLMs + OCR.
• Build and optimise RAG (Retrieval-Augmented Generation) workflows for structured credit data including chunking strategies, embedding models, and vector store management.
• Collaborate with the architect to evaluate and integrate AI coding assistants (GitHub Copilot, Cursor, Claude Code), automated PR review bots, and test-case generation pipelines into existing Forgejo/Jenkins CI/CD workflows.
• Develop prompt engineering standards, fine-tuning strategies, and model evaluation frameworks aligned with internal SDLC quality gates.
• Own MLOps infrastructure: model versioning, A/B testing, drift detection, and observability on GCP (Vertex AI, Cloud Run, BigQuery ML).
• Drive data analytics initiatives — building LLM-powered reporting layers, anomaly detection, and insights extraction from loan performance data.
• Establish AI engineering best practices, conduct technical spike reviews, and mentor junior/intern engineers.
• Stay ahead of LLM landscape shifts (open-weight models, multimodal, agents) and present actionable adoption proposals to leadership.

Skills

AI code review

LLM-assisted development

MLOps

GCP Vertex AI

CI/CD automation

Jenkins

Tools

GitHub Copilot

Cursor

Claude Code

Forgejo

BigQuery

Job description

FinacPlus is a Great Place to Work® Certified organization — a recognition of our people first culture, collaborative environment, and focus on professional growth. We provide high-end virtual business process and technology services to leading global clients across finance, banking, mortgage, research, and data services. At FinacPlus, you’ll work with world-class talent, cutting-edge technology, and international stakeholders — while still enjoying the openness, agility, and career visibility of a fast-growing company.

About Our Client — Toorak Capital Partners

You will be part of a dedicated engineering team supporting Toorak Capital Partners, a leading U.S.-based integrated correspondent lending platform that funds residential, multifamily, and mixed-use real estate loans across the U.S. and U.K. Headquartered in Summit, New Jersey, Toorak’s leadership team brings deep expertise across real estate lending, capital markets, securitization, asset management, and credit. To date, Toorak-funded projects have renovated or stabilized housing for 9,000+ families — averaging 500+ families every month. This is a rare opportunity to build mission-critical, cloud-based mortgage technology platforms that directly support one of the most respected players in the global mortgage finance industry.

ROLE SUMMARY

You will design, build, and operate AI-powered systems that process mortgage and structured credit documents at scale — spanning classification, extraction, splitting, and compliance tagging. In parallel, you will partner with our engineering architect to embed AI-assisted developer productivity tools (code generation, automated PR review, test coverage analysis) across a 50-engineer org. This is a dual-track role: product-facing AI systems and platform engineering for AI adoption.

KEY RESPONSIBILITIES

• Design and maintain intelligent document pipelines for mortgage artefacts — loan packages, appraisals, title reports, closing disclosures — covering classification, splitting, field extraction, and audit tagging using LLMs + OCR.
• Build and optimise RAG (Retrieval-Augmented Generation) workflows for structured credit data including chunking strategies, embedding models, and vector store management.
• Collaborate with the architect to evaluate and integrate AI coding assistants (GitHub Copilot, Cursor, Claude Code), automated PR review bots, and test-case generation pipelines into existing Forgejo/Jenkins CI/CD workflows.
• Develop prompt engineering standards, fine-tuning strategies, and model evaluation frameworks aligned with internal SDLC quality gates.
• Own MLOps infrastructure: model versioning, A/B testing, drift detection, and observability on GCP (Vertex AI, Cloud Run, BigQuery ML).
• Drive data analytics initiatives — building LLM-powered reporting layers, anomaly detection, and insights extraction from loan performance data.
• Establish AI engineering best practices, conduct technical spike reviews, and mentor junior/intern engineers.
• Stay ahead of LLM landscape shifts (open-weight models, multimodal, agents) and present actionable adoption proposals to leadership.

REQUIRED SKILLS

Developer Productivity AI: AI Code Review Automation Test Generation (LLM-assisted) MCP / Claude Code SonarQube Integration Forgejo Webhooks

EXPECTATIONS AT LEVEL

• Independently scope and deliver AI features end-to-end: from data ingestion to deployed API to monitoring dashboard.
• Lead design reviews for AI components; document architectural decisions with clear trade-off analysis.
• Demonstrate measurable impact — extraction accuracy, latency SLOs, developer time saved — with quantified baselines.
• Proactively identify risks in LLM outputs (hallucination, PII leakage, compliance edge cases) and implement mitigation guardrails.
• Contribute to org-wide AI adoption playbook and internal knowledge base.

NICE TO HAVE

Prior exposure to mortgage, structured credit, or financial services document workflows. Experience with agentic AI frameworks (LangGraph, AutoGen). Published models on Hugging Face or open-source AI contributions. Familiarity with Apache Superset or BI tooling for analytics layers.

We are considering only Immediate Joiners or candidates who are currently serving their notice period with 15 days or less remaining.