AI & RISK ENGINEER

ANUJ JANDHYALA

I build |

Computer Science student at Mahindra École Centrale specializing in risk-aware agentic software. Bridging the gap between complex stacks like LangGraph and business value.

ANUJ_AGENT_V1_ONLINE
import langgraph, fastapi, xgboost
# Initializing risk analysis agent...
agent = LangGraphAgent(risk_aware=True)
agent.evaluate(loans=150000)
>> Optimization: CACHING_ACTIVE [Haiku]
>> Decision: CREDIT_RISK_AGENT_OK (98.4%)

The Engineering Thesis

I specialize in building intelligent, autonomous systems engineered for reliability, safety, and real business return on investment.

Currently a third-year Computer Science Engineering student at Mahindra École Centrale, I focus on the intersection of AI systems and risk-aware software development. My path is unique: I'm not just writing agentic workflows; I am an IRM Level 1 Global ERM Certified student. This means I analyze software architectures through the lens of enterprise risk, cost efficiency, and business scalability.

In an era where LLM calls can get prohibitively expensive and unpredictable, I build robust architectures using tools like LangGraph and FastAPI. I leverage techniques like in-memory caching, token management, and RAG-based context compression to keep performance high and operational costs low.

What I Am Pursuing:

Bridging Technology & Risk: Transitioning agentic AI systems from academic toys into enterprise-grade, cost-optimized, and risk-managed business solutions that provide verifiable ROI.

Agentic Systems

Designing workflows using LangGraph and LangChain that make systems self-correcting, stateful, and capable of complex multi-agent cooperation.

Enterprise Risk (ERM)

Certified in Enterprise Risk Management (IRM Level 1). Applying risk analysis matrices directly to AI decision boundaries and token budgets.

Performant Backend

Architecting low-latency, scalable APIs with FastAPI and Express, coupled with modern databases (PostgreSQL, Pinecone, Firebase).

Selected Engineering Artifacts

April 2026 - May 2026

Credit Risk RAG Agent

A production-grade agentic AI system for assessing credit risk. Evaluated on the Lending Club dataset containing 150,000+ loans. Implements self-correcting routing nodes to evaluate borrower viability, risk parameters, and financial ratios.

  • Optimized LLM API costs via in-memory caching and Claude Haiku.
  • Implemented RAG-based context compression to keep token budgets low.
  • Utilized LangGraph for advanced stateful agent control flow and XGBoost for baseline tabular validation.
Feb 2026 - April 2026

CampusConnect

A real-time, context-aware full-stack web dashboard that connects university students and staff. It processes location data, schedule revisions, and announcements, rendering immediate updates through efficient state management.

  • Developed with Express, PostgreSQL backend, and a React frontend.
  • Enabled automated context switching and periodic background updates.
  • Configured secure authentication and database synchronization via Firebase and Postman testing scripts.

Technical Architecture

AI & RAG Systems

LangGraph (State & Router Control) 90%
LangChain (LCEL) 85%
RAG (Context Compression & Dense Retrieval) 88%
Claude AI (Haiku/Sonnet System Design) 85%

Backend & Databases

FastAPI & REST APIs 88%
Pinecone & FAISS (Vector DBs) 80%
Node.js / Express 82%
PostgreSQL & Firebase 80%

Tools & Infrastructure

Git & Version Control 90%
Postman (API Testing) 85%
Railway & Netlify (Cloud Deployment) 80%

Verified Credentials

IRM Level 1 Global ERM

Institute of Risk Management, UK

Enterprise Risk Management

Claude Code in Action

Anthropic Developer Ecosystem

Claude 101

Anthropic Ecosystem Training

Education & Milestones

Aug 2023 - Present Hyderabad

Bachelor of Technology (Computer Science)

Mahindra École Centrale (Mahindra University)

Focusing on algorithms, artificial intelligence pipelines, and data systems. Developing software with a core emphasis on system design patterns, concurrency, and risk mitigation models.

Relevant Coursework: Design & Analysis of Algorithms Big Data Analytics Reinforcement Learning
Completed May 2023 Hyderabad

Intermediate (Senior Secondary School)

Delhi Public School

Achieved a score of 82% overall. Actively engaged in student representation and leadership.

Leadership & Achievements:
  • Member of the Student Council
  • Appointed as the school's Literary Prefect (2023)
  • Placed third in the All India NIFT 2023 writing competition

Anuj AI Agent Console

Type commands in the console below to interact directly with the agent. Type help to list commands.

anuj-agent@system:~ v1.0.0-online
Initializing Anuj-Agent-v1.0.0 node in secure sandbox...
Connecting to Pinecone Vector Index [skills_and_history_idx]... Connected.
Agent engine fully initialized. RAG state: ACTIVE.

Welcome to the portfolio CLI. To ask the AI Agent about Anuj's qualifications, type commands or choose queries below.
Type help for a list of available CLI commands.

anuj-agent:~$
Quick Queries:

Initiate Connection

Let's collaborate

Whether you are looking to build agentic workflows, deploy risk-aware backend architecture, or discuss enterprise AI pipelines—my door is always open.

Location

NCC Urban Gardenia, Gachibowli Hyderabad, 500032