Vishal Kumar

CTO

Noida, Uttar Pradesh, India11 yrs 4 mos experience
Most Likely To SwitchAI Enabled

Key Highlights

  • 10+ years in software engineering with AI focus
  • Patent holder in AI/software systems
  • Multi-cloud certified across GCP and AWS
Stackforce AI infers this person is a Senior AI Architect specializing in multi-agent systems and cloud-based AI solutions.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Model Context Protocol (mcp)Cloud ComputingDatabasesJavascriptData AnalysisMachine Learning

Other Skills

A2APython (Programming Language)Vue.jsFull-Stack DevelopmentKubernetesTeam LeadershipSoftware ArchitectureAgile MethodologiesADKLangChainGoogle ADKA2A ProtocolRAGHallucination MitigationVector DBs

About

With 10+ years in software engineering and a deep recent focus on multi-agent architectures, I bridge the gap between cutting-edge AI research and enterprise-grade engineering. My work spans designing LLM-powered agent systems, implementing Model Context Protocol (MCP) servers, building Agent-to-Agent (A2A) communication frameworks, implementing RAG pipelines with grounding and hallucination mitigation strategies, and deploying these systems on GCP, AWS, and Azure What I've built: Agentic Application— An enterprise multi-agent AI platform using Google ADK v1.0, orchestrating specialized sub-agents via A2A protocol v0.2. Features RAG pipelines, a three-tier agent memory system built on mem0 (short-term, long-term, hindsight reflection loops), agent & MCP tool authentication via JWT + KeyCloak, and Gemini 2.5 Flash as the reasoning backbone. My edge: I hold patent filings in AI/software systems — I don't just implement ideas, I originate them. Multi-cloud certified across GCP (Associate Cloud Engineer) and AWS (Solutions Architect Associate) — I deploy what I design. Deep security expertise across the AI stack — from KeyCloak-based identity federation and MCP tool-level authorization to cryptographic controls (payload signing, token encryption, secret rotation) for agent-to-agent communication. Infrastructure-as-code and DevOps depth — Terraform for full cloud provisioning, GitLab CI/CD pipelines, and Kubernetes — I own the full deployment lifecycle of every AI system I build. Currently at Thales as Senior Technical Lead, driving AI adoption across critical systems. Open to AI Architect, Principal AI Engineer, and AI Platform roles where I can shape the agentic AI strategy at scale. Core stack: Python • Google ADK • MCP • A2A Protocol • RAG • Hallucination Mitigation • Vector DBs • mem0 Agent Memory • LangChain • LangGraph • LangSmith • MLflow • PyTorch • Hugging Face • Vertex AI • Gemini / OpenAI / Claude APIs • Agent & MCP Tool Authentication • KeyCloak • Cryptography • FastAPI • Kubernetes • GCP • AWS • Azure • Terraform • GitLab CI/CD • SQL

Experience

11 yrs 4 mos
Total Experience
1 yr 10 mos
Average Tenure
2 yrs
Current Experience

Thales

Senior Technical Lead

Jun 2024Present · 2 yrs · Noida · Hybrid

  • Leading AI Engineering Team for the development of high-impact software solutions using A2A, MCP, LLM, Agent Orchestrations etc.
  • Collaborate with cross-functional teams to design and optimize database schemas, resulting in significant performance enhancements.
  • Drive the adoption of best practices in software development, DevOps, cloud computing, and AI/ML.
  • Architecting and leading development of an enterprise multi-agent AI platform using Google Agent Development Kit (ADK), orchestrating domain-specific AI sub-agents via A2A protocol v0.2 for IT operations, cloud cost management, and workforce development workflows.
  • Designed and implemented MCP (Model Context Protocol) server integrations enabling LLM-powered agents to consume structured enterprise data and APIs as tools.
  • Implemented agent memory architecture using mem0 — covering short-term (in-session context), long-term (persistent user/domain knowledge via mem0's memory layer).
  • Designed agent authentication and MCP tool authentication framework — securing agent-to-agent calls and MCP server tool invocations using JWT bearer tokens, scoped credentials, and KeyCloak-backed identity management.
  • Implemented cryptographic controls for inter-agent message integrity and secure credential handling — including token signing, payload encryption, and secret rotation strategies.
  • Implementing prompt engineering patterns, context management, and fallback strategies for production-grade reliability.
  • Designed LLM evaluation pipelines using LangSmith — covering faithfulness, groundedness, and answer relevance — enabling systematic quality assurance for agentic AI workflows in production.
  • Implemented observability and monitoring for agentic workflows — covering agent trace logging, tool call latency, token consumption, and failure analysis for continuous production health monitoring.
  • Leveraged Vertex AI on GCP for managed model serving and pipeline orchestration.
Artificial Intelligence (AI)Model Context Protocol (MCP)A2APython (Programming Language)Vue.jsJavaScript+7

Oracle

Senior Software Engineer

May 2021Jun 2024 · 3 yrs 1 mo · Bangalore Urban, Karnataka, India

  • Design, develop, troubleshoot and debug software programs for databases, applications, tools, networks etc.
  • Assist in defining and developing software for tasks associated with the developing, debugging or designing of software applications or operating systems. Provide technical leadership to other software developers. Specify, design and implement modest changes to existing software architecture to meet changing needs.
DatabasesVueJavaScriptTypeScriptCloud ComputingObject-Oriented Programming (OOP)+9

Orange business services

Senior Software Engineer

Jul 2020May 2021 · 10 mos · Gurgaon, Haryana, India

VueJavaScriptCloud ComputingObject-Oriented Programming (OOP)Python (Programming Language)Kubernetes

Mahindra comviva

Senior Engineer

Jul 2018Jul 2020 · 2 yrs · Bengaluru Area, India

  • EDA, Processing, cleansing, and verifying the integrity of data used for analysis
  • Python libraries such as NumPy, SciPy, Pandas, scikit-learn, matplotlib, Seaborn and BeautifulSoup.
  • Understanding of machine learning techniques and algorithms, such as k- NN, Naive Bayes, SVM, Decision Forests, Logistic Regression etc.
  • HyperTuning and Model Tuning using K-Fold and Cross-val-score
  • LabelEncoding and OneHotEncoding, pipeline, JSON
  • Doing ad-hoc analysis and presenting results in a clear manner
  • Develop insights from a vast amount of data, graph visualization
  • Knowledge of modeling and ML (Supervised ML, Unsupervised ML).
  • Configuring nginx webserver on windows, linux and Ubuntu.
  • Basic knowledge of AWS cloud EC2,VPC etc for deploying the models into production.
  • Python Flask server, JSON, JS, HTML for model testing.
  • Good scripting and programming skills in Python
  • Experienced in deploying and integrating solutions at customer sites. Wide area of experience includes working on Linux, MySQL, Postgres DB, SHell Scripting etc.
JavaScriptCloud ComputingObject-Oriented Programming (OOP)Python (Programming Language)KubernetesData Analysis+1

Ericsson

Engineer

Jun 2016Jul 2018 · 2 yrs 1 mo · Bengaluru Area, India

  • Prepare data sets for analysis and interpretation.
  • Conceive, design and build scalable analytics solutions, using data wrangling, analysis and visualization
  • Troubleshoot, debug and upgrade existing systems.
  • Configuring nginx webserver on windows, linux and Ubuntu.
  • Develop scripts, tools and application
  • Developing Shell and Python scripts for automation of user administration and daily checks of servers.

Tech mahindra

Associate Software Engineer

Jan 2015May 2016 · 1 yr 4 mos · Pune Area, India

  •  Develop scripts, tools and application
  •  Developing Shell and Python scripts for automation of user administration and daily checks of servers.
  •  Supporting the operations team in case of any issue in production.
  •  Working on OSS Tools such as OSS-RC, Assure Now and OAM.

Education

NERIST

Bachelor of Technology (B.Tech.) — Computer Science and engineering

Jan 2010Jan 2014

Kendriya Vidyalaya ONGC Agartala, Tripura

class Xth

Jan 2004Jan 2008

Stackforce found 100+ more professionals with Artificial Intelligence (ai) & Model Context Protocol (mcp)

Explore similar profiles based on matching skills and experience