Kuldeep Bishnoi

About Hypertest: No-code API-testing tool that auto-generates end-to-end tests from real network traffic so engineering teams catch bugs before release. -------------------------------------------------------------------------------- At Hypertest, I developed the core software feature that onboarded 7 enterprise clients and generated $126K in its first year. I also made our software 30% smaller for faster deployment and improved code efficiency by 33%. My quick resolution of critical bugs for key clients like Nium and other ensured their satisfaction and contract renewals

• Built Hypertest's plug-and-play bytecode instrumentation core (HLD→LLD→Prod), securing 7 enterprise clients and $126K ARR in its first 12 months.
• Optimized the hypertest library size by eliminating unused artifacts cutting the SDK footprint by 30% (from 20MB) thus speeding up deployment pipelines
• Refactored projects by extracting common utilities into a shared module, improving code reuse and maintainability, reducing duplicate code by 33% (approx)
• Automated multi-version compatibility tests for Redisson, Jedis, OkHttp3, and AMQP, flagging version conflicts pre-deployment and caught 0-day incompatibilities before deployment.
• Championed client success for Nium, Fareye, Loop Health, and Sodel, fixed most critical bugs in less than 24 h and earned "exceptional support" and locking in every client's annual renewal.

About Fabhotel: Asset-light chain of standardized budget hotels with 600 + properties in 50 + Indian cities , offering reliable , affordable stays to travelers. ---------------------------------------------------------------------------------As a Backend Developer at Fabhotels, I drastically sped up a data gathering tool from 40 seconds to 6 seconds. I engineered a discreet API scraper and scaled systems to handle 10 times more traffic by optimizing memory use. I also made flight searches 300ms faster for over 5,000 monthly users.

• Developed a selenium scraper featuring dynamic IPs; reduced latency from 40 seconds to 6 seconds by addressing warm-up issue & proxy latency; deployed the solution on GKE with Selenium Grid.
• Engineered an undetectable API scraper in Go that bypasses Akamai WAF, featuring JA3 fingerprint spoofing, dynamic IPs, and custom header ordering.
• Scaled system to handle 10x traffic by optimizing memory usage via Datadog monitoring and HeapDump analysis
• Reduced Flight Search API latency by 300ms/call by consolidating 3 REST calls into 1, improving user experience for 5k+ monthly users
• Reduced latency by 7% by replacing REST with gRPC for inter-microservice communication.

About Aimlarger : Aimlarger is a premier provider of advance technology services, specialising in tailored A. I. Solutions and web development solutions. ----------------------------------------------------------------------------------I took full ownership of backend system development, from initial design to deployment, for a service-based company delivering web solution. A key success was spearheading the Aadhaar Hospital website project—I managed client onboarding, led backend development, and ensured its successful launch.

• Independently engineered backend systems and managed deployment for a service-based company delivering advanced web solutions, working with a team of 4 people.
• Designed and built the Aadhaar Hospital website from scratch, managing client onboarding, backend development, and deployment to ensure a successful project delivery.
• Contributed to core technical development while effectively coordinating with team to guarantee smooth project execution and timely delivery.

During my Microsoft internship, I designed and built a facial recognition attendance system. The system achieved 97.3% accuracy in tests and was successfully deployed on Microsoft Azure.

• Architected an Springboot attendance recording application using Deep Learning and gRPC for communication.
• Utilized face-recognition library and achieved 97.3% test accuracy on Labelled Faces in the Wild Dataset.
• Successfully deployed both Python application and Springboot backend on Microsoft Azure VM for reliability.

During my TCS ION internship, I developed an image classification model for ford motor client and achieved 98.8% accuracy on the test dataset.

• Developed an image classification model using TensorFlow and employed Transfer Learning with VGG16 model
• Successfully addressed overfitting challenges using dropout and regularizer & achieved 98.8% accuracy on test set