You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: src/components/styles.js
+26-28Lines changed: 26 additions & 28 deletions
Original file line number
Diff line number
Diff line change
@@ -27,27 +27,27 @@ const experiences = [
27
27
icon: tl,
28
28
iconBg: "#E6DEDD",
29
29
date: "Jul. 2024 - Present",
30
-
points: [
31
-
"- Developed features handling timeout and error conditions, improving transaction reliability.",
32
-
"- Optimized Ingenico’s payment solution, reducing lag and improving uptime by 15% across 10 functionalities.",
33
-
"- Refactored legacy code using SOLID principles, reducing technical debt by 20% and maintenance time by 25%.",
34
-
"- Migrated 20+ projects from SVN to Bitbucket, automating scripts to streamline deployment.",
35
-
"- Collaborated with a Scrum team using Jira, achieving an average sprint completion rate of 95%.",
36
-
],
30
+
points: [
31
+
"- Implemented C++ design patterns (Factory, Observer, Iterator, Strategy) to modularize payment-terminal firmware, improving maintainability and enabling rapid feature rollouts within biweekly sprints.",
32
+
"- Developed certificate-expiry detection logic in C++, cross-checking X.509 certificates against the system clock at startup; terminals now refuse expired credentials and display a 'Certificate Expired' prompt, protecting transactions for thousands of merchants.",
33
+
"- Refactored legacy code paths with modern C++ best practices (lambda expressions, std::unique_ptr, std::function maps), reducing authorization routine execution time by 15% and eliminating critical memory leaks.",
34
+
"- Authored Python scripts to migrate over 50 SVN repositories to Bitbucket, automating branch mapping and preserving full commit history, reducing migration effort by 40%.",
35
+
"- Triaged and resolved high-severity customer issues (terminal reboots, system crashes) by leading 10+ root-cause analysis sessions—examining logs, reproducing firmware faults, and updating configuration guides—to reduce mean time to resolution by 30%.",
36
+
"- Mentored three junior engineers through internal DevOps and advanced C++ certification programs, co-leading weekly hands-on workshops to share best practices in CI/CD pipelines and modern C++ features.",
"- Optimized NextVision, an application for interpreting seismic data using parallel programming.",
47
-
"- Analyzed existing algorithms in NextVision.",
48
-
"- Created a Docker environment to reduce setup time by 50%, for seamless deployment.",
49
-
"- Developed a hybrid CPU-GPU algorithm, Improving execution time by 5.39x.",
50
-
"- Tested and benchmarked to confirm performance enhancements.",
46
+
"- Accelerated BFS graph processing by 5.39× on 44 GB seismic datasets by refactoring NextVision’s C++ code to use cugraph::bfs with concurrent multi-source searches and optimized CUDA kernel queues via OpenACC pragmas on an NVIDIA A100 GPU.",
47
+
"- Reduced NextVision environment setup time by 50% (from 40 minutes to 20 minutes) by designing a Docker container pre-installing CUDA 12, cuGraph 0.21, Thrust, and Python 3.10, ensuring reproducible HPC onboarding and eliminating configuration drift.",
48
+
"- Validated algorithmic correctness via byte-for-byte patch comparison, confirming that GPU-accelerated BFS outputs matched CPU results across 120 test graphs (1 GB–44 GB) and maintained 100% fidelity in cluster assignments.",
49
+
"- Collaborated with geophysicists and GIS analysts to identify maximal independent vertex subsets—automating multi‑seed BFS launches—and co-authored an EAGE conference paper that demonstrated a 5× improvement in interpretive throughput.",
50
+
"- Benchmarked FastAPSP versus BFS workloads, projecting a further 3× reduction in compute complexity on large sparse graphs; documented findings to guide future cuGraph integrations once API support becomes available.",
"- Developed a Cloud-Based HPC Training Platform with Advanced Features and Enhanced User Experience.",
74
-
"- Designed and implemented dashboards for six distinct roles, each with customized KPIs and departmental metrics.",
75
-
"- Secured communication between backend and frontend using JWT authentication.",
76
-
"- Created features for courseprogress tracking and automated certification issuance.",
72
+
"- Built CI/CD pipelines with GitHub Actions—accelerating releases by 80%—and containerized microservices with Docker and Kubernetes for scalable HPC/AI deployments.",
73
+
"- Architected AWS (EC2, S3, IAM) infrastructure and implemented Prometheus/Grafana monitoring for 99.99% uptime and proactive alerting on performance issues.",
74
+
"- Migrated infrastructure to Terraform as code, improving deployment repeatability and reducing environment drift.",
75
+
"- Automated course-progress tracking and certificate issuance workflows, reducing manual tasks for trainers by 40%.",
77
76
],
78
77
},
79
78
{
80
79
title: "Full-Stack Developer Intern",
81
80
company_name: "Djagora FABLAB",
82
81
icon: dg,
83
82
iconBg: "#E6DEDD",
84
-
date: "Jul. 2022 - Aug. 2022",
83
+
date: "Jul. 2022 - Sep. 2022",
85
84
points: [
86
-
"- Developed a mobile application for patient monitoring, used by over 50 patients during initial testing, improving healthcare accessibility and efficiency.",
87
-
"- Implemented real-time visualization features for tracking patient conditions, with updates displayed within a 1-second delay.",
88
-
"- Designed an alert system capable of sending notifications in under 5 seconds, reducing response time for critical emergencies by 40%.",
89
-
85
+
"- Developed a Flutter/Dart mobile app for patient monitoring, delivering sub‑second vital‑sign updates for 50+ users.",
86
+
"- Engineered an alert subsystem to send notifications within 5 seconds of critical condition detection, reducing response times by 40%.",
87
+
"- Integrated MongoDB with a Python/Flask backend to securely store clinical data and power RESTful API endpoints for seamless data exchange.",
88
+
"- Collaborated with healthcare professionals to gather requirements and optimize UX, achieving over 50 active users during initial testing.",
0 commit comments