Shivam Mishra
Santa Clara, California, United States
4K followers
500+ connections
About
Senior Software Engineer | Platform & Infrastructure Specialist
Engineering platforms…
Activity
4K followers
-
Shivam Mishra shared thisEvery quarter at NVIDIA, we get two Global Free Days where the whole company truly disconnects — no meetings, no messages, just time to recharge. They’re typically attached to a weekend, so you can turn them into a longer break with family or friends. This time, we made a quick trip to LA and took our two-year-old to Disneyland. Watching his face light up all day was the best reminder of why this time off matters. Grateful to Jensen and NVIDIA for intentionally building in space to rest, reset, and be present with the people who matter most. #NVIDIA #NVIDIALife #FreeDays #FamilyTime
-
Shivam Mishra shared thisMulti-agent workflows are starting to look a lot like the new microservices. For years we’ve been breaking systems down into small, focused services with clear contracts, ownership, and SLAs. Now we’re starting to do the same thing with AI: instead of one giant “do everything” assistant, we’re seeing specialized agents for code review, testing, security, migrations, deployments, incident response, and more — all coordinated like a distributed system. The interesting part (to me) isn’t the “wow, it writes code” moment anymore. It’s questions like: - How do you design an architecture where a review agent, a test agent, a security agent, and a deployment agent all hand work off to each other safely? What does observability look like when agents are reading repos, calling APIs, updating configs, and opening PRs on their own? - How do we apply the same ideas we already know — idempotency, retries, backoff, circuit breakers, least privilege — to agent workflows instead of just microservices? - If you squint a bit, an “agent mesh” starts to look a lot like a service mesh: routing, policies, limits, audit trails, and a control plane that keeps the whole thing understandable. As engineers and architects, this is where it gets fun. The job shifts from “Can AI do X?” to “How do we design the system so that a group of narrow, reliable agents can do X, Y, and Z together without surprising us?” I don’t think agents replace microservices — I think they become another layer in the same architecture. Curious how others are thinking about this: Are you already experimenting with multiple specialized agents in your workflow (review, tests, CI/CD, security), or is everything still going through one big “AI assistant” today?
-
Shivam Mishra posted thisOver the last year, AI coding agents have evolved from “interesting prototypes” to tools that can reliably open PRs, run tests, and handle real engineering workflows. One of the biggest blockers for serious teams, though, has stayed the same: security and environment boundaries. It’s hard to adopt powerful agents if it means shipping sensitive code and execution to someone else’s cloud. That’s why Cursor’s new capability — cloud agents that run on your own infrastructure — feels like an important step forward. Instead of executing in a vendor-controlled environment, agents now run inside your network: your VPCs, your build machines, your security perimeter. Your code, artifacts, and secrets stay on your machines, while Cursor provides the orchestration layer and developer experience on top. From an architect’s perspective, this unlocks a few meaningful shifts: • Security and compliance: Teams in regulated domains (finance, healthcare, gov, enterprise SaaS) can adopt agentic workflows without relaxing their “code stays in our environment” policies. • Realistic execution environments: Agents run against the same dependencies, internal services, feature flags, and networks your engineers use every day, which reduces the “it worked in a sandbox, failed in prod” gap. • Deeper automation: You can hand off end-to-end flows — regression runs, refactors, dependency upgrades, security patches, CI/CD triage — to agents that behave more like well-governed service accounts than one-off scripts. What I like about this direction is that it aligns AI with how we already design systems: clear boundaries, least-privilege access, and strong observability around what’s running where. AI agents stop being mysterious black boxes “somewhere in the cloud” and start looking more like a new layer in our existing architecture. We’re still early in the journey of bringing agents into real-world engineering environments, but this feels like one of those infrastructure changes that will look obvious in hindsight. If your team could run coding agents inside your own infra, what’s the first workflow you’d delegate — tests, refactors, security fixes, CI/CD debugging, or something else?
-
Shivam Mishra shared thisWhat a week at GTC 2026 in San Jose. Today was the last day, and I finally had time to walk the expo floor and really take in what people are building with AI — not just the big keynotes, but the real work happening behind the scenes with partners, vendors, and customers. From infrastructure providers like CoreWeave and Nebius offering NVIDIA GPUs at cloud scale, to booths from AWS, Microsoft, and many others showcasing how they’re integrating agentic AI into their services, it was incredible to see the full stack come together — hardware, software, and applications. What excited me most was seeing how teams are taking NVIDIA’s platforms and building real-world solutions on top: from model training and inference to domain-specific apps and tools that are already in production. As a tech enthusiast (and someone lucky enough to work with this ecosystem day to day), it’s inspiring to see where the industry is headed. I’m leaving GTC feeling energized about the pace of innovation and the quality of work happening across the AI ecosystem. Excited to see what this holds for the future.
-
Shivam Mishra posted this🚀 One of the most compelling ideas from GTC wasn’t just a new platform announcement — it was a new way to think about the future of computing. With NVIDIA’s Vera Rubin Space‑1 and its broader space computing vision, we may be seeing the early foundation of a world where AI infrastructure is no longer confined to Earth-based data centers. For decades, the model has been familiar: generate data remotely, transmit it back, process it centrally, and respond after the fact. But if compute can run closer to where data is actually created — in orbit — that changes the equation in a meaningful way. The implications are significant. For Earth observation, communications, climate monitoring, defense, and autonomous space systems, on-orbit AI could reduce latency, ease bandwidth constraints, and enable faster decisions in environments where speed and autonomy matter most. What makes this especially interesting is the problem it aims to solve. We are creating more data at the edge than traditional architectures can efficiently move and process in time, and space-based compute introduces a new way to rethink that bottleneck. Of course, bold infrastructure shifts come with real challenges. Radiation, thermal constraints, power efficiency, reliability, and the economics of orbital data centers will all shape how quickly this vision becomes practical at scale. Still, the broader idea is hard to ignore. We’ve seen computing evolve from centralized systems to cloud, from cloud to edge, and now potentially from Earth-bound infrastructure to space-native computing. We may still be early, but the direction is compelling. Excited to see what this holds for the future. Curious how others see this: Is space-based computing the next major infrastructure shift, or will it remain focused on a few high-value use cases?
-
Shivam Mishra posted this🚀 How Nvidia’s Nemo Claw Is Reshaping Developer Workflows After 10+ years in software engineering — much of it spent refining how we build, test, and ship code — I’m genuinely excited about tools that don’t just assist developers but collaborate with us across the full lifecycle. Enter Nemo Claw: Nvidia’s enterprise-grade evolution of OpenClaw, designed to tackle real-world business and engineering challenges. Think of it as an AI agent that integrates deeply into your workflow: Code review: It analyzes changes, suggests fixes, and flags potential issues — not just syntax, but logic and architecture concerns. - Test automation: Generates unit/integration tests, runs them, and even updates suites as features evolve. - CI/CD intelligence: Monitors pipelines, diagnoses failures (flaky tests, resource bottlenecks), and proposes fixes — reducing MTTR. - Ticket-to-action: Connects to JIRA, reads bug/feature tickets (e.g., “Pneumocystis minor bug”), drafts MRs for review, and assigns owners contextually. This isn’t about replacing engineers — it’s about elevating our role. By automating repetitive, high-friction tasks, Nemo Claw frees us to focus on design, innovation, and solving harder problems. For teams drowning in technical delivery overhead, this could be a force multiplier. The future of enterprise development isn’t just faster — it’s smarter, more responsive, and deeply integrated. 💭 How is your team preparing to adopt AI agents like Nemo Claw in your workflows? What’s one process you’d offload first — code reviews, test writing, pipeline debugging, or something else?
-
Shivam Mishra posted thisOver the last few months I’ve quietly rebuilt my day‑to‑day workflow around two tools: Claude Code in the terminal and Cursor as my main IDE. Together, they’ve made “agentic AI” feel a lot less like a buzzword and a lot more like having an extra junior dev (or two) on the team that never sleeps. My mental model is simple: Cursor is where I think with the code, and Claude Code is where I delegate the grunt work. In Cursor, I’m iterating fast on ideas, spiking out designs, and letting the built‑in agents touch multiple files at once when I’m refactoring or wiring up a new end‑to‑end slice. When the change gets big enough that I’d usually open a fresh branch and start grinding, I hand it off to Claude Code in the terminal and let it plan, edit, run tests, and propose commits while I stay in review mode. The “latest” shift for me hasn’t been one big feature drop, it’s the way these tools now work at the repository level instead of as autocomplete in a single file. Claude Code can walk an entire codebase, build a plan, and execute it step by step; Cursor’s agents can take an idea and push it through multiple files while still keeping me in the loop in the editor. That combination is what finally made agentic AI feel production‑ready for real systems instead of a neat demo. Does it magically write perfect software? No. I still spend a lot of time reading diffs, tightening up designs, and saying “no” when an agent reaches for the wrong abstraction. But as someone who’s spent the last decade building scalable, event‑driven systems, I’m now spending less energy on boilerplate and more on architecture, invariants, and the parts of the job that actually require judgement. If you’re a developer who already lives in a terminal and a modern editor, my honest take is: pair Claude Code with Cursor for a few real features, not toy examples, and see how it changes the way you think about your own role in the stack.
-
Shivam Mishra posted thisAnthropic just proved AI can build a 100K-line compiler autonomously. Here's what it means for your career. They scaled Claude Code into a fully autonomous "Agent Team" that built a 100,000-line Rust-based C compiler. In two weeks. For $20,000 in API credits. And it successfully compiled Linux 6.9 and Doom. This isn't a demo. This is a proof of concept for autonomous engineering departments. 𝗪𝗵𝗮𝘁 𝗵𝗮𝗽𝗽𝗲𝗻𝗲𝗱: 16 parallel Claude Code instances ran in Docker containers. Some wrote core logic. Others acted as "Reviewers," "Documentarians," and "Performance Optimizers." They managed their own Git flow with a locking system. They resolved merge conflicts. They maintained a shared task list. When something failed, they automatically analyzed logs and retried. The researchers essentially walked away and let the agents work. 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: We're moving past "AI as a co-pilot." We're entering "AI as a department." The economics are clear: • $20K for two weeks of autonomous development • That's a fraction of a specialized team's cost • The agents worked 24/7 without breaks But here's what caught my attention: The agents eventually hit a "complexity ceiling." Fixing one bug would break another. They needed human intervention to break the cycle. This reveals the future role of software engineers: We won't be writing code. We'll be architecting the test harnesses. We'll be designing the constraints. We'll be breaking the infinite loops. We'll be making the trade-offs that agents can't. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗶𝘀 𝗺𝗲𝗮𝗻𝘀 𝗳𝗼𝗿 𝘆𝗼𝘂: The skills that matter: → System architecture and design → Defining clear constraints and boundaries → Building robust test frameworks → Understanding business context → Making judgment calls on trade-offs The skills that matter less: → Writing boilerplate → Routine debugging → Syntax memorization → Manual code reviews 𝗛𝗼𝘄 𝘁𝗼 𝗽𝗿𝗲𝗽𝗮𝗿𝗲: 1. **Elevate your thinking** Move from "how do I implement this?" to "how do I architect this?" 2. **Master test design** The agents will write code. You'll write the tests that keep them honest. 3. **Understand constraints** Learn to define boundaries, guardrails, and success criteria clearly. 4. **Develop domain expertise** Agents can't make business decisions. You can. 5. **Embrace orchestration** Your role shifts from coder to conductor of an AI orchestra. The future is coming, whether you like it or not. The question isn't "will AI replace developers?" The question is "will you be ready to architect the systems that AI builds?" Are you preparing for this shift? --- #SoftwareEngineering #AI #FutureOfWork #TechLeadership #ClaudeCode #AutonomousAgents #NVIDIA
-
Shivam Mishra posted this"ERROR: Payment failed for user." Great. - Which user? - Which payment? - Why did it fail? This log is of no help. Here's the difference: Bad Log: ERROR Payment failed Good log: { "error": "card_declined", "user_id": "usr_456", "trace_id": "abc123", "duration_ms": 234 } Now I can: ✓ Query all failures for this user ✓ Jump to the distributed trace ✓ See exactly what went wrong ✓ Correlate with other events The rules: 1. Structured JSON (not strings) 2. Always include trace_id 3. Add business context 4. Use log levels correctly 5. Never log secrets What's the worst log message you've seen in production? #Logging #Observability #DevOps 💡 Complete guide in comments
-
Shivam Mishra liked thisShivam Mishra liked thisAs I close an incredibly meaningful chapter, I wanted to share that March 27th marked my last day at AWS. Leaving Amazon on a high was always important to me—and I’m grateful to say I’ve been able to do just that. Over the past 8+ years, the longest tenure in my career so far, this journey has been nothing short of transformative—both professionally and personally. At AWS, I had the privilege of working alongside some of the most talented, driven, and intelligent individuals across the globe. From partnering with diverse business leaders across multiple Amazon businesses to collaborating with exceptional Talent Acquisition leaders, every experience has contributed to shaping my perspective and strengthening my craft. Amazon has been a masterclass in navigating complexity—whether it was scaling during high-growth phases, adapting through re-orgs, or pivoting during unprecedented times like COVID. The culture of innovation and being ahead of the curve in technology has constantly pushed me to stay curious, agile, and forward-thinking. What truly stands out are Amazon’s Leadership Principles—they go beyond being workplace tenets and have become a way of life. They’ve deeply influenced how I think, make decisions, and solve problems, both at work and beyond. I am deeply thankful to the incredible AWS Business leaders, TA leadership, mentors, and colleagues who invested in me, challenged me, and supported my growth throughout this journey. The relationships and friendships built here are ones I will always cherish. Thank you, AWS, for shaping me into a stronger, more resilient professional. I carry these experiences forward with immense gratitude and pride. As I look ahead, I’m excited about new beginnings and the next chapter in my journey—I’ll be sharing more very soon. To all Amazonians—thank you for the memories, the learning, and the lifelong connections. #Gratitude #NewBeginnings #AWS #Leadership #CareerJourney #Amazon Bandita Acharya Finuala Hattori Josh Procter Gary Qin Anil Warrier Stephanie Walker Jay Shankar Aswathy Pillai Ponnamma Bellathanda Nikhita Rai K Tejasvi Kochar Anal Saha Jackson Varghese Nisha Sunil Pankaja Sridhar Muhammad Ridwan Arlene Koh Stella Pedziwiatr Yatindra Sharma Krithi Marla Savari Gotadki Samta Chopra Dipika Ramachandran Ron Richard Ramirez Aayushi Bajpai Sajid Akram Guddi Thapa Mridul Kalita Soundarya Rao Manjunath Jiddimanii Ramcharan Teja Sripathi Risa Ogawa Vicki Low Fibha M. Dylan White Veda Nadgouda Stuti Punn Sampada Masur Palkar Iori Kudo
-
Shivam Mishra liked thisShivam Mishra liked thisWe are pleased to announce that RNSIT students won the BGS Hackathon and received a cash prize of Rs. 50,000. 💐
-
Shivam Mishra liked thisShivam Mishra liked thisI wrote five markdown files on product principles this week and put them in every Aurigo repo. Kevin Koenig our CTO, Shashi Kumar Balu our VP Product and Sanat Kumar Mohapatra, our Director of Product Engineering got them deployed across the codebase. Not a wiki. Not a Confluence page. Markdown files at the root of the repo, next to the code. Each one covers who the customer is, what the product philosophy is, what we never compromise on, and what our AI architecture looks like. When an AI coding assistant opens a file in your repo, it has no context. It does not know who your customer is, what your moat is, or what you never compromise on. So it writes perfectly functional code that slowly drifts from your product philosophy in ways that are hard to see and harder to reverse. The product principles file at the root of the repo is the answer. One shared file covering who we serve, why we exist, and what good product looks like at Aurigo. Four product files covering each of our product lines. Every engineer reads the shared file when they join. Every AI coding assistant reads both before touching the code. Three things I learned writing them. Writing it down forces decisions you have been avoiding. You cannot write “our moat is twenty years of domain data and deep domain expertise encoded into our agents” without confronting whether your team actually builds that way every day. A principles file is not documentation. Documentation describes what the system does. A principles file describes why every decision was made and what you never compromise on. Those belong in different places. And the file from the founder lands differently than a wiki page from a product manager. Engineers know when the philosophy comes from the top. One gets read. The other gets ignored. If you are building an AI-native company and you have not written this file yet, do it today. Not for your engineers. For the AI coding assistants working alongside them who have no other way to understand what you are building and why. We build software that builds the world. Every line of code should know that.
Experience
-
-
-
-
-
-
-
-
-
Education
-
Visvesvaraya Technological University
-
-
Projects:
Hand Gesture Recognition using OpenCV-
The goal of this project was to design and build a man-machine interface using a video camera to interpret one-handed gestures (plus others for additional keyboard and mouse control).
The keyboard and mouse are currently the main interfaces between humans and computers.
Humans communicate mainly by vision and sound, therefore a man-machine interface would be more intuitive. No physical contact to the input devices…Projects:
Hand Gesture Recognition using OpenCV-
The goal of this project was to design and build a man-machine interface using a video camera to interpret one-handed gestures (plus others for additional keyboard and mouse control).
The keyboard and mouse are currently the main interfaces between humans and computers.
Humans communicate mainly by vision and sound, therefore a man-machine interface would be more intuitive. No physical contact to the input devices required !!!
IDE - Visual Studio | Languages/Libraries - C++ OpenCV
3D Maze Game-
3D-Maze is a simple maze constructed in 3D, using shaders based OpenGL. The maze is constructed using meshes and scaling factors with a background using cube mapping. Navigation through the maze is a first person camera, where the user can rotate the camera using the mouse and move using the keyboard functions.
IDE : Visual studio | Languages: C++, OpenGL, GLSL -
-
-
-
-
-
Licenses & Certifications
Patents
-
Application management platform for hyper converged cloud infrastructures
Filed US 2022 039525
Courses
-
C Programming
-
-
C#.Net
-
-
Computer Graphics and Visualization using OpenGL
-
-
Computer Networks
-
-
Data Structures
-
-
Database Management Systems (DBMS)
-
-
Design and Analysis of Algorithms
-
-
Graph Theory
-
-
Java
-
-
Logic Design
-
-
Microprocessors
-
-
Object Oriented Programming Using C++
-
-
Operating Systems
-
-
Web Programming
-
Projects
-
Cloud Infrastructure Efficiency Engineering (CIE) – IPP Core Infra, NVIDIA
Led design and development of a cloud-scale observability and diagnostics platform for NVIDIA’s centralized build and test infrastructure, covering more than 5,000 GPU nodes with plans to scale to tens of thousands.
Role and responsibilities
- Drove architecture, implementation, and rollout of a system agent, Java API services, and backend data pipelines used by SREs to maintain infrastructure stability across the company.
- Partnered with SRE teams to convert recurring…Led design and development of a cloud-scale observability and diagnostics platform for NVIDIA’s centralized build and test infrastructure, covering more than 5,000 GPU nodes with plans to scale to tens of thousands.
Role and responsibilities
- Drove architecture, implementation, and rollout of a system agent, Java API services, and backend data pipelines used by SREs to maintain infrastructure stability across the company.
- Partnered with SRE teams to convert recurring infrastructure issues into automated detection, faster root-cause analysis, and self-service diagnostics.
System capabilities
- Built a lightweight Golang agent that runs as a background service on each node, collecting rich CPU/GPU, hardware, and software metadata on configurable schedules and pushing it into Elasticsearch for fast indexed search.
- Developed a Java Spring Boot API layer that lets SREs query and compare nodes at scale, including “diff” operations between a single node and a pool, or across arbitrary node sets, to expose configuration and state differences.
- Exposed APIs for before/after snapshots around test runs so teams can diff system state over time and correlate changes with test failures.
Impact on reliability
- Enabled rapid identification of issues such as mismatched BIOS and driver versions, incompatible package stacks, abnormal GPU temperatures, and problematic CPU/GPU family combinations (for example, specific AMD CPU and Blackwell GPU pairings), as well as subtle version and configuration drifts across fleets.
- Reduced debug time from multi-day investigations to hours or less for many infrastructure-related test failures by providing a single, queryable source of truth for node state, history, and cross-node comparisons.
Tech stack
Technologies: Golang agent, Java Spring Boot APIs, PostgreSQL, Elasticsearch, internal data platforms, and modern observability tooling for logging and metrics. -
Centralized Notification Service – NGC / Internal Platforms (NVIDIA)
-
Built a centralized Notification Service that abstracts the complexity of event-based communication for internal products, providing a single, reusable platform for customized, multi-channel notifications. Microservices across the organization could onboard easily and start emitting notifications without reimplementing delivery logic or preference management.
The service enabled end users to subscribe to specific events and configure how they wanted to be notified (email, Slack, group…Built a centralized Notification Service that abstracts the complexity of event-based communication for internal products, providing a single, reusable platform for customized, multi-channel notifications. Microservices across the organization could onboard easily and start emitting notifications without reimplementing delivery logic or preference management.
The service enabled end users to subscribe to specific events and configure how they wanted to be notified (email, Slack, group chat, and other internal channels), with flexible templates to curate the content and reports sent for each notification type. Notification delivery was automated and triggered in real time as events occurred in upstream systems.
Contributed to the architecture, design, and end-to-end implementation, including CI/CD, deployment, and ongoing enhancements, while leading a team of two engineers to deliver the service into production for multiple products. This platform reduced duplication of effort across teams and ensured a consistent, reliable notification experience for internal users. -
CICD as a service
-
Served as the principal architect and lead developer for an enterprise-grade CI/CD-as-a-Service platform that automated software delivery pipelines across NVIDIA’s internal ecosystem. The platform enabled engineering teams to onboard applications instantly, with automated build, testing, versioning, quality certification, and multi-environment deployment—all governed through a unified workflow.
Led the end-to-end architecture, development, deployment, and production operations of the…Served as the principal architect and lead developer for an enterprise-grade CI/CD-as-a-Service platform that automated software delivery pipelines across NVIDIA’s internal ecosystem. The platform enabled engineering teams to onboard applications instantly, with automated build, testing, versioning, quality certification, and multi-environment deployment—all governed through a unified workflow.
Led the end-to-end architecture, development, deployment, and production operations of the system, ensuring high availability for mission-critical products. Played a key role in incident management and platform reliability, collaborating with cross-functional teams to maintain seamless operations and rapid recovery.
Mentored junior engineers, established best practices for scalable system design, and drove adoption of modern DevOps principles organization-wide.
The platform became a core productivity enabler, significantly accelerating software delivery cycles and ensuring consistent quality across teams.
Technologies: .NET Core, React, Redux, Node.js, PostgreSQL, Elasticsearch, Material UI, Docker, Kubernetes, Helm, Jenkins, Artifactory, OpenTelemetry -
Automated Lifecycle Manager (ALM) – NGC Cloud Group (NVIDIA)
-
I was a part of the architecture and development team of Automated Lifecycle Manager (ALM), a platform automating the complete lifecycle of NVIDIA applications deployed on on-premise Kubernetes clusters (e.g., Metropolis). ALM enabled secure and seamless release distribution, update scheduling, and version management across distributed client environments.
Designed and implemented the Kubernetes ALM Operator to detect new releases, notify connected clusters, and execute automated or…I was a part of the architecture and development team of Automated Lifecycle Manager (ALM), a platform automating the complete lifecycle of NVIDIA applications deployed on on-premise Kubernetes clusters (e.g., Metropolis). ALM enabled secure and seamless release distribution, update scheduling, and version management across distributed client environments.
Designed and implemented the Kubernetes ALM Operator to detect new releases, notify connected clusters, and execute automated or scheduled application updates. Built the integrated CI/CD pipeline handling certification, release, and deployment workflows with full traceability and compliance.
Technologies: Go, .NET Core, Kubernetes, Helm, Jenkins, React, PostgreSQL, OpenTelemetry -
Metrics [24 7.ai]
-
Designed and implemented a centralized analytics platform to surface engineering and business metrics from large-scale HDFS data to key stakeholders across the organization, including product management, executive leadership, marketing, revenue optimization, and operations teams.
Built role-based authentication and authorization to provide tailored dashboard views and metric access per stakeholder group, ensuring data governance, security, and relevance for each…Designed and implemented a centralized analytics platform to surface engineering and business metrics from large-scale HDFS data to key stakeholders across the organization, including product management, executive leadership, marketing, revenue optimization, and operations teams.
Built role-based authentication and authorization to provide tailored dashboard views and metric access per stakeholder group, ensuring data governance, security, and relevance for each audience.
Technologies: Node.js, React, Hadoop, Spark, Elasticsearch. -
End to End Test Automation Framework [24 7.ai]
-
Designed an automation framework used to validate the flow of data from the source to the big data platform.
The framework was integrated with TestNG. It took advantage of all the hooks provided by TestNG to perform operations before and after various events(test,suite,execution). It also had a retry analyzer which could be customized to re-run the tests that had failed.
The framework could capture the network requests that were made for each test and validate them. It was…Designed an automation framework used to validate the flow of data from the source to the big data platform.
The framework was integrated with TestNG. It took advantage of all the hooks provided by TestNG to perform operations before and after various events(test,suite,execution). It also had a retry analyzer which could be customized to re-run the tests that had failed.
The framework could capture the network requests that were made for each test and validate them. It was integrated with browser stack, which enabled the UI tests to be executed on OS and browsers passed at run time.
Technologies - Java, Selenium, BrowserStack, BrowserMob -
Automation Dashboard [Aurigo Software Technologies]
-
See project
Automation Dashboard is a web application that was built to track the automation progress, give visualization for different metrics like current automation status and coverage, progress made on automation each week, trigger automation for different products and have the status monitored from the dashboard and publish the test results on the dashboard.
Technologies Used :
ASP.Net MVC, MSSQL, Entity Framework, HTML, CSS, Bootstrap, Javascript -
Automation Framework - Optimus [Aurigo Software Technologies]
-
Optimus is an automation framework built to write automated tests for Masterworks. The framework defines a set of standards that must be followed to make the tests robust and maintainable.
Highlights of the framework are-
1. Modular in nature - Separate module for Executor (responsible for executing tests), Common Utilities (Reporting, DB Access, Emails, helper classes for xml and excel, etc) and Test Scripts.
In order to execute tests from another team, just plug in the tests…Optimus is an automation framework built to write automated tests for Masterworks. The framework defines a set of standards that must be followed to make the tests robust and maintainable.
Highlights of the framework are-
1. Modular in nature - Separate module for Executor (responsible for executing tests), Common Utilities (Reporting, DB Access, Emails, helper classes for xml and excel, etc) and Test Scripts.
In order to execute tests from another team, just plug in the tests dll and its available for execution.
2.Configurable - all the configurations are kept separate in xml files. Settings like build on which automation has to run, emails, reports, publishing report on the dashboards, what type of run (smokes, full pass, etc ), browser to be used for test, priorities of tests to be executed, etc can be configured just by modifying few values.
3.Reporting - There are multiple types of reports that are generated for each automation run - HTML, xml and text.
4.The framework supports UI Testing and API Testing.
5.The framework supports parallel Execution of tests module wise. Feature for executing individual tests in parallel in under progress.
We have a separate dashboard for tracking progress of automation for each team in the organization. -
Online Teachers Appraisal System (OTAS) [RNSIT]
-
OTAS is a web Application which is used for the performance evaluation of the professors in the college. The data was collected from the students at the end of each semester and reports were generated which were reviewed by the HOD of all the departments and the higher management.
It had a role based authentication mechanism in which the same application was catered to different users like students, teachers, Head of Departments and the Principal.
The pages that were rendered were…OTAS is a web Application which is used for the performance evaluation of the professors in the college. The data was collected from the students at the end of each semester and reports were generated which were reviewed by the HOD of all the departments and the higher management.
It had a role based authentication mechanism in which the same application was catered to different users like students, teachers, Head of Departments and the Principal.
The pages that were rendered were based on their roles.
The Website was hosted on a local server in the college
Technologies Used :
ASP.NET MVC, MSSql, Bootstrap, JQuery, HTML, CSS, Microsoft SSRS for reporting
and Entity Framework -
Project Open house Panorama (PROP) [RNSIT]
-
It is a state level inter collegiate Project Open house Panorama hosted by the Department of Computer Science and Engineering RNSIT. It facilitates learning and career enhancement research and exploring of ideas in all categories of Information Technology. It gives and opportunity for companies to identify individuals with exceptional capabilities to develop functional projects.
Technologies used for the website -
HTML, CSS, JQuery, MySQL, PHPOther creators -
Honors & Awards
-
Young Innovator of the Year
Aurigo Software Technologies
-
Spot Award
Aurigo Software Technologies
Languages
-
English
Native or bilingual proficiency
-
Hindi
Native or bilingual proficiency
-
Marathi
Professional working proficiency
-
Maithili
Full professional proficiency
Recommendations received
5 people have recommended Shivam
Join now to viewOther similar profiles
Explore more posts
-
enreap
JFrog Fly is redefining how engineering teams manage artifacts in an AI-driven world. As software delivery becomes faster, distributed, and increasingly automated by AI, traditional artifact management can’t keep up. JFrog Fly introduces a smarter, globally scalable way to store, replicate, and deliver artifacts—built for modern engineering velocity. 🔗 Read the full blog to explore what this shift means for your DevOps pipeline. Link - https://lnkd.in/dUK_UFd8 #JFrog #AI #DevOps #SoftwareDelivery #CloudNative
-
SYNNQ
Most SDKs help you build apps. SYNNQ Pulse SDK builds infrastructure. It is now live. Not for chatbots. Not for quick demos. This SDK enables AI systems to train across institutions, devices, and regulatory domains. You can now: 🟡 Train LLMs and ML models in federated environments 🟡 Deploy multi-client training across nodes and GPU types 🟡 Keep data local while synchronising global model updates 🟡 Run edge learning without orchestration rewrites 🟡 Apply system-level compliance with GDPR, EU AI Act, and local policy constraints It supports real-time training in finance, healthcare, mobility, defense, and any domain with sensors, data, and distributed compute. Built for teams building AI infrastructure. Contact us for access and support. — SYNNQ Pulse European Superintelligence Infrastructure
-
Career Ahead
Waymo has resumed its robo-taxi service in San Francisco after a power outage. This highlights the importance of resilience in tech-driven industries. Consider how disruptions can impact your career and develop strategies to adapt and thrive. #CareerAdvice #TechIndustry #Resilience #AutonomousVehicles
-
Redis
LiteLLM is an open-source proxy that connects your app to LLMs through one unified interface. Paired with Redis, it gives AI and ML teams a simple yet powerful way to unify access to LLMs, accelerate response times, and make AI apps real-time. Redis handles performance, memory, and data coordination that modern AI apps demand, while LiteLLM handles abstraction and routing. Our AI Product Marketing Manager Rini Vasan dives into: ▶️ What LiteLLM is ▶️ Why it works well with Redis ▶️ How LiteLLM and Redis work together Learn how to scale your LLM gateway with LiteLLM & Redis: https://lnkd.in/gqsTebF7
-
QuantEdge Technologies
We just published our first technical research note. "Low-Latency Event-Driven Architecture for Real-Time Options Execution on IBKR" It covers how we eliminated compounding feed lag at SPY market open — from a growing 45-second backlog to stable sub-second bar sealing — through: → Queue-first callback offloading (Python GIL problem) → Per-venue DOM isolation across ARCA, ISLAND, IEX → Event-time binning across 6 independent intervals → Single source of truth for all technical indicators No signals. No alpha. Pure infrastructure. This is the kind of work that happens before a single trade fires — and the kind that determines whether your system survives the opening bell. 📄 Download: https://lnkd.in/g-XJsYwG #AlgoTrading #OptionsTrading #MarketMicrostructure #SystematicTrading #Python #QuantitativeFinance #TradingInfrastructure
-
Bassirah Technology
AI systems struggle with accuracy and memory. Redis solves this with semantic caching, vector search, and scalable memory enabling agentic RAG at enterprise scale. As the official Redis distributor in Egypt, Bassirah is proud to help enterprises bring these advanced AI capabilities into production. #AI #DataManagement #DataIntegration #AIsolutions #TechSolutions
-
io.net
The harsh reality: 73% of distributed systems fail to scale beyond initial deployment. Is your AI startup heading for a compute crisis? ML workloads need 40x more resources than traditional applications, but most teams are using infrastructure strategies that break at production scale. The companies that make it past this bottleneck understand something critical about distributed architecture that the 73% who fail completely miss. It's not about adding more machines. It's about three specific design decisions that determine whether your system scales or collapses under load. Link in comments for the full breakdown of what separates the 27% who succeed.
1 Comment -
Emerj Artificial Intelligence Research
"Pull all siloed AI projects into a one-page chart. Focus on the ones aligned with strategic goals, pause the rest, assign owners, remove roadblocks, and make clear what business value each delivers." – Julian Tang, AI Champion, BlackRock In this Pure Storage–sponsored episode of the #AIinBusinessPodcast, Julian Tang joins Emerj Artificial Intelligence Research Editorial Director Matthew DeMello to break down how enterprises can operationalize AI, align projects with strategy, and accelerate measurable outcomes. Retail and CPG leaders can uncover actionable frameworks to integrate AI into workflows, prioritize initiatives, and unlock ROI with greater clarity and speed. Listen to the full episode to discover actionable strategies for aligning AI projects with business goals and accelerating real results: https://lnkd.in/dxxHiDGU #EnterpriseAI #AIGovernance #AIWorkflow
1 Comment -
ZETIC
On-device AI becomes powerful only when it works reliably on real hardware. In this Melange Walkthrough, Yeonseok Kim and Sun-Q Kim break down how the full on-device pipeline works, from model selection and quantization to real-device benchmarking and SDK deployment. Select. Benchmark. Deploy. This is how production-ready on-device AI should operate. Watch the full walkthrough below. #OnDeviceAI #EdgeAI #MobileAI #AIInfrastructure #ZETIC #Melange
-
SogetiLabs
How Large Language Models Are Making Software Greener AI is quietly driving sustainability. Large Language Models (LLMs) now go beyond generating text—they’re optimizing code. By detecting inefficiencies and refining logic, they cut CPU cycles, reduce energy use, and lower costs. Cleaner code means greener software. At scale, these improvements shrink carbon footprints, turning LLM-assisted refactoring into an environmental win. Explore how AI is shaping a future where innovation meets sustainability. https://lnkd.in/dfwy_Yhq #AI #SustainableTech #GreenSoftware #CodeOptimization #Innovation Md Siddiqur RAHMAN, Mouna Ben Mabrouk Ha Nhi NGO Ginel Dorleon Azade Fotouhi José Carlos Rosales Núñez Ines BEN KRAIEM
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named Shivam Mishra
7194 others named Shivam Mishra are on LinkedIn
See others named Shivam Mishra