Indira Keerthana Pulla

Indira Keerthana Pulla

Full-Stack & AI Systems Developer | Crafting High-Performance Backends & Agents


About

Hey! πŸ‘‹ I’m Indira, a developer passionate about bridging complex backend systems with modern AI agent architectures. What I build & focus on: AI Agents & Developer Tools – Context-aware tools, agentic workflows, and MCP architectures. Backend Systems – High-concurrency APIs, distributed logic, and scalable cloud setups. Open Source & Tech Communities – Actively exploring system design, LLM evaluations, and builder ecosystems. Always up for discussing agent orchestration, backend optimization, or cool open-source projects! Drop a message or let’s connect at DevFest.

Skills
TypeScript
JavaScript
Python
React
Node.js
Vertex AI
MongoDB
Projects

DetectThreatAI

An AI-assisted email forensic investigation platform for security analysts - parses .eml files, analyzes headers/SMTP relay paths, verifies SPF/DKIM/DMARC authentication, extracts IOCs (IPs, domains, URLs, attachments), pulls threat intelligence (IP reputation, geolocation, ASN, DNS/RDAP, VPN/TOR indicators), correlates entities into an investigation graph, and generates an AI-powered threat classification with confidence scoring and recommended actions. Also handles case/campaign management, evidence integrity (SHA-256 hashing, provenance tracking), and exports forensic PDF.

React
TypeScript
Vite
Tailwind CSS. Backend β€” Python
FastAPI
Pydantic. Database β€” PostgreSQL/Supabase + SQLite. AI β€” Groq API. Visualization β€” Leaflet (geolocation maps)
Cytoscape (investigation graph). Reporting β€” ReportLab (PDF generation). Threat intel sources β€” AbuseIPDB
VirusTotal
DNS
RDAP. Deployment target: Render (backend) + Vercel (frontend).

AgentHiveAI

An agentic logistics support system that resolves shipment exceptions (damaged packages, failed deliveries, missing documentation, bad addresses, delayed shipments) β€” the agent investigates via carrier tracking tools, plans a resolution, runs a risk assessment, and either auto-executes low-risk actions or routes high-risk ones to a human-approval queue with a full reasoning trail shown in the UI.

Python backend
Streamlit
FastAPI + Uvicorn
Groq API running Llama 3.3 70B
Supabase
custom tool-calling agent loop
risk-policy engine for the approval gating.

Trade-Flow

Zerodha-inspired stock trading platform with real-time price updates via WebSockets; includes custom-built components like a support page replica.

MERN stack (MongoDB
Express
React
Node.js)
WebSockets
JavaScript
Education

B.Tech in Computer Science

Ace Engineering College Β· 8.94

Sept 2024 to Apr 2028
Certifications

Claude on Google Cloud

Anthropic Education

Jun 2026

Introduction to Cybersecurity

Cisco Networking Academy

Jan 2026
DevFestHYDERABAD

Attendee Β· Oct 24, 2026

About DevFest Hyderabad