Harsh Raj Srivastav profile photo 1 on Tal

Harsh Raj Srivastav

Founding Engineer @ClickPe (YC W23) | Backend + Cloud + AI Infrastructure | Building Scalable Systems 0→1 | AI Automation

Bengaluru

4+ years of work experience

Rolaa AI
Apr 20251 yr 6 mosPresent

Bengaluru

Building rolaa.ai — an AI meeting intelligence platform that auto-joins calls, transcribes meetings, and sends automated follow-ups via email and WhatsApp.
ClickPe (YC W23)
May 20233 yrs 5 mosPresent

Bengaluru

Initial engineering hire at Clickpe YC W23 fintech startup, building technical infrastructure from 0→1 while wearing multiple hats across engineering, product, and operations. Led architecture decisions enabling rapid growth and enterprise partnerships. Core Platform & Infrastructure:   • Implemented Loan Management System using Python/AWS Lambda improving operational efficiency by 25%, directly enabling first enterprise deals   • Developed custom Apache Fineract wrapper with Alembic/SQLAlchemy for seamless database migrations, minimizing manual intervention   • Designed microservices architecture on AWS (VPC, ALB, Auto Scaling, ECR, CloudFront) ensuring high availability for enterprise operations   • Deployed Temporal on EKS/ECS with Go-based orchestration service, boosting scalability and fault tolerance AI Automation & Voice Systems:   • Built voice calling agents with TTS/STT, RAG/NLP using self-hosted LiveKit, achieving 30-40% higher lead conversion, reduced loan drop-off, and sub-900ms multilingual call support   • Created document verification system with AWS Rekognition/OpenCV achieving 95% accuracy, eliminating manual review bottlenecks   • Automated marketing/operations/HR workflows with n8n agentic systems, reducing overhead by 70%   • Enhanced credit underwriting by migrating Camunda v7→v8, Drools improving scalability and reliability Business Impact:   • Integrated 15+ lending partners and financial service providers, reducing vendor onboarding time by 60%   • Built technical hiring process and scaled team from 1 to 5 engineers   • Integrated CIBIL credit scoring, bank analysis, documents generations AWS SES workflows driving 40% more automation Technical Stack: Python, Go, AWS (Lambda, EKS, ECS, VPC), Apache Fineract, Temporal, Camunda, n8n, LiveKit, Kubernetes
Aug 20236 mosJan 2024

At college, I developed QuizMate, a quiz management application that addresses the challenges students and educators face with college quizzes typically conducted through Google Forms. This platform enhances the quiz experience with key features: 1. Question Management: View, create, update, and delete questions easily. 2. Topic and Difficulty Filtering: Filter questions by predefined topics and difficulty levels. 3. User Profiles: Create profiles, verify emails, and manage profile pictures using Cloudinary. 4. Session Management: Create private quiz sessions, invite participants, and track unique session IDs. 5. Blog Features: Engage with posts, comments, and likes. 6. Google Vertex AI Integration: Generate hints for quiz questions. 7. Security: APIs secured with Spring Security, requiring JWT tokens for sensitive actions. This project involved collaboration with various students, particularly in the frontend domain, where we built the website using React.js and developed a Flutter application. Through QuizMate, I improved user engagement and streamlined quiz management, enhancing the overall learning experience while honing my skills in Java Spring Boot, PostgreSQL, and AWS.
Liveasy
Dec 20227 mosJun 2023

Remote

Integrated components and security filters in Java Spring Boot, utilizing PostgreSQL and implementing JWT-based security measures to enhance application protection, resulting in a 40% reduction in unauthorized access incidents . Automated deployment processes by configuring builds with Maven and Docker, integrating with a GitHub Actions CI/CD pipeline featuring a self-hosted runner, which decreased deployment times by 30% and ensured seamless deployment on EC2. Integrated AWS WAF and other security components, effectively managing the secure flow of requests through load balancer target groups, enhancing application security and achieving an uptime of 99.9% during the project period. Implemented Grafana, Loki, and Prometheus for the Spring Boot application, providing a visual dashboard for logs and system monitoring, which improved observability and performance tracking of the application, leading to a 25% increase in response time efficiency.
Oct 20228 mosMay 2023

Kanpur, Uttar Pradesh, India

Harsh’s AI Stack

langchain
langchain
langgraph
langgraph
n8n
n8n
ollama
ollama

Education

Harcourt Butler Technical University

Bachelor of Technology - BTech, Information Technology

Harcourt Butler Technical University

Rani Laxmi Bai Memorial School (R.L.B.)

Intermediate, Science

Rani Laxmi Bai Memorial School (R.L.B.)

20191 yr2020

Join Harsh and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Harsh Raj Srivastav on Tal

Headline

Founding Engineer @ClickPe (YC W23) | Backend + Cloud + AI Infrastructure | Building Scalable Systems 0→1 | AI Automation

About

I'm a passionate founding engineer who builds scalable backend systems and AI automation infrastructure for early-stage startups. At ClickPe, I've transformed operational efficiency through intelligent automation, reducing manual processes by 40% while building production-grade fintech solutions. What I Build: • Backend & Cloud Infrastructure: Java, Python, Go with AWS (Lambda, EKS, ECS, VPC, Auto Scaling) • AI-Powered Automation: Voice calling agents with TTS/STT, RAG/NLP pipelines, automated workflows using Temporal and n8n • Serverless Microservices: Designed fault-tolerant architecture reducing MTTR by 80% and debugging time significantly while ensuring proper request handling • Financial Systems: Built Loan Management System improving operational efficiency by 25%, integrated Apache Fineract with seamless database migrations My Approach: Instead of grinding LeetCode problems, I focus on solving real-world challenges through POCs that become production features. I spend weekends connecting with startup founders to understand their pain points and build solutions that scale from 0→1. Problem-Solving Track Record: • Designed serverless microservices architecture orchestrated with Temporal workflows deployed on EKS, reducing MTTR by 80% and debugging time significantly while ensuring proper request handling and fault tolerance • Integrated 10+ open-source tools (Apache Fineract, Temporal, Camunda 8, LiveKit, n8n, Keycloak, Grafana) into production-ready solutions • Created voice calling agents increasing lead conversion rates by 30-40% • Built enterprise-grade monitoring solution achieving 99.9% system visibility • Enhanced credit underwriting workflows migrating from Camunda v7 to v8 • 500+ DSA problems solved, but more importantly - countless startup challenges solved I'm passionate about: • Building rapid prototypes in startup environments • Contributing to live products that create real impact • Automating tedious tasks with AI to unlock human creativity • Working with small teams making big impacts (speedboat over Titanic mentality) Perfect for: YC startups, remote-first companies, and founders who need someone to build their technical foundation from day one. Let's connect if you're building something that matters! 🚀

Location

Bengaluru

AI stack

langchain

langgraph

n8n

ollama

Work experience

Building Rolaa ai at Rolaa AI, Apr 2025 - Present · 1 yr 6 mos, Bengaluru, Building rolaa.ai — an AI meeting intelligence platform that auto-joins calls, transcribes meetings, and sends automated follow-ups via email and WhatsApp.

Founding Engineer at ClickPe (YC W23), May 2023 - Present · 3 yrs 5 mos, Bengaluru, Initial engineering hire at Clickpe YC W23 fintech startup, building technical infrastructure from 0→1 while wearing multiple hats across engineering, product, and operations. Led architecture decisions enabling rapid growth and enterprise partnerships. Core Platform & Infrastructure: • Implemented Loan Management System using Python/AWS Lambda improving operational efficiency by 25%, directly enabling first enterprise deals • Developed custom Apache Fineract wrapper with Alembic/SQLAlchemy for seamless database migrations, minimizing manual intervention • Designed microservices architecture on AWS (VPC, ALB, Auto Scaling, ECR, CloudFront) ensuring high availability for enterprise operations • Deployed Temporal on EKS/ECS with Go-based orchestration service, boosting scalability and fault tolerance AI Automation & Voice Systems: • Built voice calling agents with TTS/STT, RAG/NLP using self-hosted LiveKit, achieving 30-40% higher lead conversion, reduced loan drop-off, and sub-900ms multilingual call support • Created document verification system with AWS Rekognition/OpenCV achieving 95% accuracy, eliminating manual review bottlenecks • Automated marketing/operations/HR workflows with n8n agentic systems, reducing overhead by 70% • Enhanced credit underwriting by migrating Camunda v7→v8, Drools improving scalability and reliability Business Impact: • Integrated 15+ lending partners and financial service providers, reducing vendor onboarding time by 60% • Built technical hiring process and scaled team from 1 to 5 engineers • Integrated CIBIL credit scoring, bank analysis, documents generations AWS SES workflows driving 40% more automation Technical Stack: Python, Go, AWS (Lambda, EKS, ECS, VPC), Apache Fineract, Temporal, Camunda, n8n, LiveKit, Kubernetes

Developing Quizmate at Quizmate, Aug 2023 - Jan 2024 · 6 mos, At college, I developed QuizMate, a quiz management application that addresses the challenges students and educators face with college quizzes typically conducted through Google Forms. This platform enhances the quiz experience with key features: 1. Question Management: View, create, update, and delete questions easily. 2. Topic and Difficulty Filtering: Filter questions by predefined topics and difficulty levels. 3. User Profiles: Create profiles, verify emails, and manage profile pictures using Cloudinary. 4. Session Management: Create private quiz sessions, invite participants, and track unique session IDs. 5. Blog Features: Engage with posts, comments, and likes. 6. Google Vertex AI Integration: Generate hints for quiz questions. 7. Security: APIs secured with Spring Security, requiring JWT tokens for sensitive actions. This project involved collaboration with various students, particularly in the frontend domain, where we built the website using React.js and developed a Flutter application. Through QuizMate, I improved user engagement and streamlined quiz management, enhancing the overall learning experience while honing my skills in Java Spring Boot, PostgreSQL, and AWS.

Back End Developer Intern at Liveasy, Dec 2022 - Jun 2023 · 7 mos, Remote, Integrated components and security filters in Java Spring Boot, utilizing PostgreSQL and implementing JWT-based security measures to enhance application protection, resulting in a 40% reduction in unauthorized access incidents . Automated deployment processes by configuring builds with Maven and Docker, integrating with a GitHub Actions CI/CD pipeline featuring a self-hosted runner, which decreased deployment times by 30% and ensured seamless deployment on EC2. Integrated AWS WAF and other security components, effectively managing the secure flow of requests through load balancer target groups, enhancing application security and achieving an uptime of 99.9% during the project period. Implemented Grafana, Loki, and Prometheus for the Spring Boot application, providing a visual dashboard for logs and system monitoring, which improved observability and performance tracking of the application, leading to a 25% increase in response time efficiency.

Building Bthames at Bthames, Oct 2022 - May 2023 · 8 mos, Kanpur, Uttar Pradesh, India

Education

Bachelor of Technology - BTech, Information Technology, Harcourt Butler Technical University

Intermediate, Science, Rani Laxmi Bai Memorial School (R.L.B.)

Public links

github: https://github.com/Harsh-Srivastav123