AI Agents Engineer

Modussecure; Bengaluru; On-site; INR 1,500,000 - 2,500,000; Full time

Job summary

Modussecure, a leading fintech AI company located in Bengaluru, is seeking a professional with expertise in AI for financial risk. The role involves designing multi-agent systems, fine-tuning LLMs, and deploying secure AI solutions for major financial institutions. Candidates should have 2-4 years of relevant experience, particularly with agent frameworks and backend services. Join a selective team where autonomy and innovation lead to impactful AI collaboration with India's largest banks and fintechs.

Qualifications

• 2-4 years of total experience in AI.
• 6-12 months of professional experience with AI agents.
• Strong understanding of financial risk, AML, and fraud.

Responsibilities

• Design and deploy multi-agent orchestration frameworks for financial risk workflows.
• Fine-tune LLMs using financial datasets.
• Build secure and compliant AI systems for banks.

Skills

LangChain / LangGraph

AutoGen Custom orchestration

LoRA / QLoRA fine‑tuning

Python (FastAPI / Flask)

PostgreSQL / Redis

Job description

Build the multi-agent intelligence layer for financial risk at India's most ambitious fintech AI company.

HSR Layout, Bengaluru Full-Time · In-Person Total Experience: 2-4 years 6-12 months professional AI Agents experience FinTech · Risk · Fraud

The Role

We build AI for Risk Management teams at financial institutions across India. If you want to architect multi-agent orchestration systems, fine‑tune LLMs for high-stakes domains, and deploy hardened on‑prem AI into the infrastructure of India's biggest banks and fintechs, this is the job.

You’ll own the full stack: from agent design and LLM fine‑tuning to backend APIs, production deployment, and client outcomes.

What You’ll Build

Multi‑Agent Orchestration

Design and deploy orchestration frameworks for financial risk workflows: fraud detection pipelines, AML case escalation, and automated credit analysis.

LLM Fine‑Tuning

Fine‑tune and align open‑source LLMs on proprietary financial datasets. Own data curation, LoRA/QLoRA training, RLHF, and evaluation.

Package and ship AI systems into air‑gapped or private‑cloud environments inside banks. Keep systems secure and audit‑ready.

RAG and Knowledge Systems

Build retrieval‑augmented pipelines grounded in regulatory documents, transaction histories, and internal knowledge bases.

Backend APIs and Services

Build and maintain backend services that power agent workflows: REST and async APIs, data pipelines, and enterprise integrations.

Evaluation and Safety

Create eval suites for financial contexts: hallucination detection, output grounding, latency benchmarks, and guardrails.

Skills We’re Looking For

Agent Frameworks

LangChain / LangGraph AutoGen Custom orchestration Tool use / function calling ReAct / chain‑of‑thought

LoRA / QLoRA fine‑tuning Hugging Face Transformers RLHF / DPO vLLM / TGI / Ollama Prompt engineering Quantization (GGUF, AWQ)

Python (FastAPI / Flask) REST and async APIs PostgreSQL / Redis Message queues (Kafka / RabbitMQ) Basic auth and API security

RAG and Data

Vector DBs (Qdrant, Weaviate, PGVector) Embeddings and chunking Structured + unstructured RAG SQL / graph DBs

Domain Bonus

Financial risk / AML / fraud Graph ML Regulatory / compliance AI

Our Culture

• Extremely selective hiring with exceptional operators who raise your bar.
• Product owners, not ticket‑closers: own features end‑to‑end and speak to clients directly.
• Full autonomy on tools, architecture, and execution.
• Research that ships into production at India's largest financial institutions.

Founded in 2025, we work with India's largest financial institutions. Built by operators from Goldman, Apple Card, Uber, NSE, KPMG, and Niyo, Modus AI ships practical risk and compliance AI into production.