U Debasish R. profile photo 1 on Tal

U Debasish R.

Building Secure Financial APIs & Infrastructure | AppSec · WAF · Detection Engineering | Senior Security Engineer @ Decentro

Odisha

6+ years of work experience

Decentro
Jul 20264 mosPresent

Bengaluru

PhonePe
Jul 20251 yrJun 2026

India

Protecting 500M+ user transactions across PhonePe's distributed fintech platform. 1.WAF Strategy & Rule Engineering: Engineered sophisticated rule sets across Cloudflare, Akamai, and F5 ASM/AFM, achieving a 45%+ reduction in malicious traffic and significantly lowering false-positive rates for production APIs. Performed continuous threat modeling and defense-in-depth tuning to block OWASP Top 10, bot, and zero-day threats. 2.In-House WAF Development: Architected and shipped a proprietary WAF leveraging Nginx Ingress Controller, Pingora, and Coraza with custom iRules-style logic; achieved a 48% improvement in bot fingerprinting accuracy and hardened the PhonePe CDN against layer-7 attacks. Deployed in Docker containers with automated CI/CD rollout. 3.AI-Augmented Alert Investigation: Designed and deployed Barbican — a production MCP (Model Context Protocol) server built on FastAPI — to expose security tools as LLM-callable functions for AI-assisted WAF alert investigation. Integrated vector embedding pipeline (all-MiniLM sentence-transformers) with FAISS cosine similarity indexing to enable semantic deduplication and intelligent alert clustering across millions of daily events, cutting L1/L2 mean-time-to-investigate by ~60%. 4.SOC Architecture & Detection Engineering: Designed an Elastic Stack-based SOC with SOAR-integrated playbooks that reduced alert noise by 30% and correlation time by 40%; integrated ML-based anomaly scoring to surface high-confidence threats from billions of daily log events. 5.Zero Trust & API Security: Established hardened API gateways and edge controls aligned with Zero Trust and defense-in-depth principles; enforced mTLS, rate limiting, and bot-scoring at the perimeter for all critical payment endpoints. 6.Performance & Scalability: Optimized WAF rule precedence and evaluation order, improving processing throughput under peak load without degrading latency SLAs.
PhonePe
Jul 20232 yrs 2 mosAug 2025

Bangalore

1.SOC Architecture & Stack Design: Served as architect of PhonePe's Security Operations Centre — designed and owned the full detection stack: Elastic Stack (ELK) as the core SIEM, Opensearch for log search and analytics, Wazuh for host-based detection and FIM, QRadar for enterprise correlation, Logstash and Rsyslog for log ingestion/normalisation, Filebeat and Vector as lightweight shippers, and SOAR playbooks for automated response — handling billions of events per day at 99.99% pipeline availability across 12,000+ systems. 2.High-Fidelity Detection Rules: Developed 500+ detection rules mapped to MITRE ATT&CK tactics across network, endpoint, and application layers; authored custom Sigma and YARA rules enabling the SOC team to identify attack patterns previously invisible to existing tooling. 3.SIEM Ingestion Pipeline Rebuild: Rebuilt the ELK + Opensearch ingestion pipeline to normalise log schemas from 20+ source systems (Wazuh agents, Filebeat, Rsyslog, Vector) enabling consistent cross-source correlation and real-time SOC dashboarding. 4.Compliance Automation & Drift Detection: Automated patch compliance workflows for 12,000+ systems using Python, Go, and Terraform; delivered hardened OS baselines (RHEL/Ubuntu, Windows) with CIS Benchmarks and implemented drift detection to auto-revert unauthorized configuration changes — improving measurable compliance posture by 60%. 5.Vulnerability Management: Owned the end-to-end vulnerability lifecycle — triaged findings from Nessus/Qualys scans, prioritized remediation based on threat intelligence, and ensured critical CVEs were resolved within defined SLAs across 12,000+ systems. 6.JA4/JA4H Bot Detection: Built a TLS fingerprinting pipeline using JA4 and JA4H signatures that reduced automated abuse attempts on payment APIs by 45% while maintaining zero friction for legitimate users.
Loginsoft
Aug 20212 yrsJul 2023

Hyderabad

Enterprise vulnerability research for Cisco, Juniper, McAfee and other Tier-1 vendors. 1.Vulnerability Research (Cisco, Juniper, McAfee): Discovered 70+ critical vulnerabilities through deep Java source code reviews and binary analysis; findings directly resulted in CVE-credited patches shipped to enterprise customers worldwide. 2.VAPT & DAST Tooling: Conducted structured penetration tests using Burp Suite Professional, OWASP ZAP, nmap, and Nikto for web app and network security assessments; identified exploitable attack chains in network device firmware and application layers across enterprise products. 3.YARA & OVAL Rule Development: Authored 6,000+ OVAL rules and purpose-built YARA signatures for malware classification and firmware threat detection; developed custom SecLang detection logic for automated scanning of proprietary binary formats against known vulnerability patterns. 4.Workflow Automation: Built Python and Go tooling to automate policy and compliance workflows, cutting manual analyst effort by 60% and enabling consistent, scalable vulnerability triage across multiple client engagements.
Jun 20201 yr 3 mosAug 2021

India

Co-founded a security services and training company delivering cybersecurity education, assessments, and secure development solutions. Highlights: > Delivered hands-on training for CEH, Security+, CCNA, Linux, helping students launch cybersecurity careers. > Secured client web-server infrastructures using ModSecurity, DDOS Deflate, Rootkit Hunter, and custom hardening scripts. > Built full-stack web applications (PHP, Python) and integrated secure CI/CD pipelines. > Provided end-to-end consulting on server security, malware protection, and controlled network architecture.
Golars Networks
Nov 20195 mosMar 2020

Hyderabad

Delivered advanced cybersecurity, networking, and cloud training programs for students and corporate clients. Key Responsibilities: > Trained in CEH, ECSA, CISA, including hands-on ethical hacking labs. > Conducted networking training on routing, switching, firewalls, and AWS cloud basics. > Delivered Linux administration training, including server security, permissions, and user management.

U’s AI Stack

claude code
claude code
claude
claude
gemini
gemini
github copilot
github copilot
n8n
n8n
notebooklm
notebooklm

Education

Bachelor of Technology - BTech, Electrical, Electronic and Communications Engineering Technology/Technician

Gandhi Institute of Technological Advancements (GITA), Bhubaneswar

20164 yrs2020

hse

Saraswati sishu vidya mandir - bhubaneswar

200410 yrs2014

Join U and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

U Debasish R. on Tal

Headline

Building Secure Financial APIs & Infrastructure | AppSec · WAF · Detection Engineering | Senior Security Engineer @ Decentro

About

I build security systems at a scale most engineers never encounter. At PhonePe, I protect 700M+ user transactions across a distributed fintech platform handling billions of events daily. That means engineering WAF rule sets across Cloudflare, Akamai, and F5 that cut malicious traffic by 45%, architecting a 500-rule detection layer on Elastic Stack that reduced alert noise by 30%, and shipping an in-house WAF on Pingora and Coraza that improved bot fingerprinting accuracy by 48% — all in production, all at scale. But the work I'm most proud of is Barbican — a production AI security orchestration platform I designed from scratch. It exposes SOC tooling as LLM-callable functions via a FastAPI-based MCP server, uses vector embeddings (all-MiniLM sentence-transformers) with FAISS cosine similarity indexing for semantic alert deduplication, and cuts L1/L2 mean-time-to-investigate by ~60%. It's what AI-augmented security actually looks like in production, not in a demo. Before PhonePe, I spent two years at Loginsoft doing deep vulnerability research for Cisco, Juniper, and McAfee — discovering 70+ critical vulnerabilities through Java source code reviews, authoring 6,000+ OVAL rules and custom YARA signatures, and contributing to CVE-credited patches shipped to enterprise customers worldwide. My expertise sits at the intersection of WAF engineering, detection engineering, cloud security, and AI-augmented defense. I work across AWS, GCP, and Azure, and I'm comfortable operating from raw packet analysis and TLS fingerprinting all the way up to compliance frameworks like PCI-DSS, ISO 27001, and SOC2. I write about WAF engineering, detection logic, and AI in security at hackervault.tech. If you're working on hard security problems at scale — or building the next generation of AI-native security tooling — I'd love to connect.

Location

Odisha

AI stack

claude code

claude

gemini

github copilot

n8n

notebooklm

Work experience

Senior Security Engineer at Decentro, Jul 2026 - Present · 4 mos, Bengaluru

Senior Infosec Engineer at PhonePe, Jul 2025 - Jun 2026 · 1 yr, India, Protecting 500M+ user transactions across PhonePe's distributed fintech platform. 1.WAF Strategy & Rule Engineering: Engineered sophisticated rule sets across Cloudflare, Akamai, and F5 ASM/AFM, achieving a 45%+ reduction in malicious traffic and significantly lowering false-positive rates for production APIs. Performed continuous threat modeling and defense-in-depth tuning to block OWASP Top 10, bot, and zero-day threats. 2.In-House WAF Development: Architected and shipped a proprietary WAF leveraging Nginx Ingress Controller, Pingora, and Coraza with custom iRules-style logic; achieved a 48% improvement in bot fingerprinting accuracy and hardened the PhonePe CDN against layer-7 attacks. Deployed in Docker containers with automated CI/CD rollout. 3.AI-Augmented Alert Investigation: Designed and deployed Barbican — a production MCP (Model Context Protocol) server built on FastAPI — to expose security tools as LLM-callable functions for AI-assisted WAF alert investigation. Integrated vector embedding pipeline (all-MiniLM sentence-transformers) with FAISS cosine similarity indexing to enable semantic deduplication and intelligent alert clustering across millions of daily events, cutting L1/L2 mean-time-to-investigate by ~60%. 4.SOC Architecture & Detection Engineering: Designed an Elastic Stack-based SOC with SOAR-integrated playbooks that reduced alert noise by 30% and correlation time by 40%; integrated ML-based anomaly scoring to surface high-confidence threats from billions of daily log events. 5.Zero Trust & API Security: Established hardened API gateways and edge controls aligned with Zero Trust and defense-in-depth principles; enforced mTLS, rate limiting, and bot-scoring at the perimeter for all critical payment endpoints. 6.Performance & Scalability: Optimized WAF rule precedence and evaluation order, improving processing throughput under peak load without degrading latency SLAs.

Infosec engineer at PhonePe, Jul 2023 - Aug 2025 · 2 yrs 2 mos, Bangalore, 1.SOC Architecture & Stack Design: Served as architect of PhonePe's Security Operations Centre — designed and owned the full detection stack: Elastic Stack (ELK) as the core SIEM, Opensearch for log search and analytics, Wazuh for host-based detection and FIM, QRadar for enterprise correlation, Logstash and Rsyslog for log ingestion/normalisation, Filebeat and Vector as lightweight shippers, and SOAR playbooks for automated response — handling billions of events per day at 99.99% pipeline availability across 12,000+ systems. 2.High-Fidelity Detection Rules: Developed 500+ detection rules mapped to MITRE ATT&CK tactics across network, endpoint, and application layers; authored custom Sigma and YARA rules enabling the SOC team to identify attack patterns previously invisible to existing tooling. 3.SIEM Ingestion Pipeline Rebuild: Rebuilt the ELK + Opensearch ingestion pipeline to normalise log schemas from 20+ source systems (Wazuh agents, Filebeat, Rsyslog, Vector) enabling consistent cross-source correlation and real-time SOC dashboarding. 4.Compliance Automation & Drift Detection: Automated patch compliance workflows for 12,000+ systems using Python, Go, and Terraform; delivered hardened OS baselines (RHEL/Ubuntu, Windows) with CIS Benchmarks and implemented drift detection to auto-revert unauthorized configuration changes — improving measurable compliance posture by 60%. 5.Vulnerability Management: Owned the end-to-end vulnerability lifecycle — triaged findings from Nessus/Qualys scans, prioritized remediation based on threat intelligence, and ensured critical CVEs were resolved within defined SLAs across 12,000+ systems. 6.JA4/JA4H Bot Detection: Built a TLS fingerprinting pipeline using JA4 and JA4H signatures that reduced automated abuse attempts on payment APIs by 45% while maintaining zero friction for legitimate users.

software engineer-security researcher at Loginsoft, Aug 2021 - Jul 2023 · 2 yrs, Hyderabad, Enterprise vulnerability research for Cisco, Juniper, McAfee and other Tier-1 vendors. 1.Vulnerability Research (Cisco, Juniper, McAfee): Discovered 70+ critical vulnerabilities through deep Java source code reviews and binary analysis; findings directly resulted in CVE-credited patches shipped to enterprise customers worldwide. 2.VAPT & DAST Tooling: Conducted structured penetration tests using Burp Suite Professional, OWASP ZAP, nmap, and Nikto for web app and network security assessments; identified exploitable attack chains in network device firmware and application layers across enterprise products. 3.YARA & OVAL Rule Development: Authored 6,000+ OVAL rules and purpose-built YARA signatures for malware classification and firmware threat detection; developed custom SecLang detection logic for automated scanning of proprietary binary formats against known vulnerability patterns. 4.Workflow Automation: Built Python and Go tooling to automate policy and compliance workflows, cutting manual analyst effort by 60% and enabling consistent, scalable vulnerability triage across multiple client engagements.

Founder at Mitigater Techsecure Pvt. Ltd., Jun 2020 - Aug 2021 · 1 yr 3 mos, India, Co-founded a security services and training company delivering cybersecurity education, assessments, and secure development solutions. Highlights: > Delivered hands-on training for CEH, Security+, CCNA, Linux, helping students launch cybersecurity careers. > Secured client web-server infrastructures using ModSecurity, DDOS Deflate, Rootkit Hunter, and custom hardening scripts. > Built full-stack web applications (PHP, Python) and integrated secure CI/CD pipelines. > Provided end-to-end consulting on server security, malware protection, and controlled network architecture.

Professional Trainer at Golars Networks, Nov 2019 - Mar 2020 · 5 mos, Hyderabad, Delivered advanced cybersecurity, networking, and cloud training programs for students and corporate clients. Key Responsibilities: > Trained in CEH, ECSA, CISA, including hands-on ethical hacking labs. > Conducted networking training on routing, switching, firewalls, and AWS cloud basics. > Delivered Linux administration training, including server security, permissions, and user management.

Education

Bachelor of Technology - BTech, Electrical, Electronic and Communications Engineering Technology/Technician, Gandhi Institute of Technological Advancements (GITA), Bhubaneswar

hse, Saraswati sishu vidya mandir - bhubaneswar

Public links

hackervault.tech: https://hackervault.tech/