I'm Nishant.

AWS Solutions Architect

DevOps Engineer

Full Stack Developer

Hiring? Don't take a résumé's word for it.

Let your AI screen me — objectively.

Paste your role, copy one self-contained prompt with my full profile baked in, and drop it into ChatGPT, Claude, or Gemini. It scores the fit — no sign-up, no install, works in any AI.

Screen me with your AI →or connect your agent →
Available now — contract & full-time, remote

Also certifiedGenerative AI Overview for Project Managers · Google Associate – G Suite Administrator · Google Analytics · Google Ads – 6 Categories · Google Digital Sales · Google Mobile Sites

services

Web Application Development

Full stack applications using React, Node.js, Python, and serverless architectures. From e-commerce platforms to AI agent interfaces, built for scale and maintainability.

Mobile Application Development

Cross-platform mobile apps with React Native and native integrations. Fintech, leasing, and e-commerce apps with secure payment gateways and real-time data.

Cloud Infrastructure and DevOps

AWS architecture with 5x certifications. Multi-region deployments, CI/CD pipelines, infrastructure as code with Terraform and CDK, and automated rollback systems.

AI and ML Integration

LLM-powered applications, RAG pipelines, voice processing systems, and AI agent platforms. Production deployments on GPU infrastructure with Triton and OpenAI APIs.

Security and Compliance

Firewall rule translation, SASE architectures, VPN systems, and cloud security automation. Check Point to AWS migration, Auth0 integration, and penetration testing labs.

Containerisation and Microservices

Docker and Kubernetes (CKAD certified) for container orchestration. Microservice architectures, SIP/VoIP systems, and multi-tenant deployment platforms.

Experience

Solutions Architect, DevOps Engineer, AI/ML Consultant

Freelance (Upwork)

2024 - Present

Update deployment systems with monitoring, rollback and multi-tenant abilities for enterprise clients.

Build email interfaces for LLM interaction and AI agent systems.

Set up Google Cloud and OpenAI infrastructure for AI Agent platforms (SAPHI).

AMIS file processing and automation for Check Point clients.

Product customizer development for e-commerce platforms.

more info

featured projects

01.

AgentOne — SAP B1 AI Automation Platform

A multi-tenant SaaS platform (sapb1.ai) that connects on-prem SAP Business One to cloud AI agents. A Go headless agent bridges the customer's SAP B1 (HANA / SQL Server) to a broker and connectors; a Python configurator self-provisions each tenant on GCP Cloud Run — Cloud SQL, buckets, schema bootstrap, and a per-tenant MCP server. ~22 microservices span an agent registry, downloadable connectors, Stripe billing, and MS Teams / WhatsApp integrations.

Python
GCP
Docker
Node.js
API
view case study →
02.

nishant-profile-mcp — AI-Readable Résumé MCP Server

An MCP (Model Context Protocol) server that exposes my professional profile — skills, projects, certifications, and experience — directly to AI recruitment agents. Ships a local-embedding semantic matcher and a 5-dimension match_role engine that scores any job description against my profile and returns a hire signal. 8 tools over stdio, fully offline (runs MiniLM embeddings locally, no API key).

Node.js
LLM
Embeddings
API
view case study →
03.

Yukt Capital — Invoice-Discounting Marketplace + MCP Server

Yukt Capital (yuktcapital.com) is a TReDS-style invoice-discounting marketplace where suppliers raise invoices on corporate buyers and investors finance them early at a fair discount. I built a companion MCP server (17 tools) that lets any MCP host — Claude Desktop, Claude Code, Gemini CLI — log in and run the full workflow in natural language: browse the marketplace, fund or bid on invoices, raise invoices, run reverse-factoring programs, and trade on the secondary market.

Node.js
Fintech
API
view case study →
04.

AQA — System of Record for Agentic Verification

A TestLink-style test-management and verification-memory platform built for the agentic coding loop, exposed over MCP (43 tools). It tracks plans, cases, dependencies, run history, and requirements-to-test traceability, with a claim/verify protocol (doer ≠ checker) and semantic regression memory: a failed run's root-cause reasoning is stored and recalled when a similar failure recurs, so agents get the prior diagnosis instead of re-deriving it. One service layer, surfaced three ways — REST API, MCP server, and CLI.

Python
API
Embeddings
view case study →
05.

l33tpwn — Cloud Pentest Training Platform (2025 Overhaul)

A ground-up rebuild of l33tPWN into a hosted, browser-based cybersecurity training platform (l33tpwn.com), in the style of HackTheBox / TryHackMe. Each student gets a private AWS lab — a Kali attack box, an optional Kali Purple blue-team/SOC box (Suricata IPS + Wazuh), and on-demand vulnerable target VMs from a 48-target catalog — all running in the browser via noVNC in ~90s. Added a leet/l33tpwn CLI (~30 verbs across student and operator faces), keyless shell access via AWS SSM Session Manager, guided walkthroughs, and EventBridge-driven idle teardown.

Python
Lambda
DynamoDB
CloudFront
Terraform
view case study →
06.

OpenClaw AI — Dream Server Open Source Contribution

Contributed to Dream Server, an open source self-hosted AI platform ('the Linux of Local AI'). Implemented OpenClaw-to-Open-WebUI integration: built HTTP API shim, LLM provider patching (OLLAMA_URL/LITELLM_KEY), Docker Compose overlay for service orchestration, and configuration layer for connecting OpenClaw's agent framework to Open WebUI's chat interface. PR accepted and personally integrated by the project maintainer into the Lemonade release branch.

Docker
Python
view case study →
07.

Autonomous AI Sales Pipeline — SAP B1 Lead Generation

Built a fully autonomous multi-agent AI sales pipeline using OpenClaw on GCP. 6-agent team: Scout (lead scraping from partner case studies), Qualifier (composite scoring with identity/event/contact dimensions), Verifier (SAP B1 confirmation), Enricher (contact discovery via Apollo/Hunter/theHarvester/Perplexica cascade), Signal Hunter 'Pulse' (intent-based scoring using Gemini/Groq for hiring/funding/complaint/migration signals), and Closer (email generation). Custom pipeline database with SQLite, real-time dashboard on port 3000, WhatsApp gateway integration, automated signal sweeps on cron. Built from scratch in 3 weeks.

Python
DockerGCP
view case study →
08.

Check Point Cloud Firewall Manager

Built a comprehensive cloud firewall management platform bridging Check Point and AWS. Created a custom SDK surpassing Check Point's own, enabling programmatic rule management across Security Groups, NACLs, Network Firewall (stateless and stateful/Suricata), and Web ACLs. Automated network feed ingestion into AWS prefix lists. Built pipeline to pull rules from Check Point, translate to intermediate format, and deploy to AWS Network Firewall with named rules and variable mapping. Deployed on GCP CloudRun with automated triggers. Set up Headscale/Tailscale VPN mesh across AWS regions for secure connectivity. 638+ hours of development.

Python
AWS
GCP CloudRun
Docker
Terraform
Boto3
view case study →
09.

Product Customizer - ColdFusion to FOSS Migration

Infrastructure rescue migrating a product customizer platform from end-of-life Adobe ColdFusion + MSSQL stack to a fully FOSS stack before ColdFusion's core support ended (Nov 2025). Migrated the application server from ColdFusion to Lucee, database from MSSQL to MySQL, containerized with Docker, and deployed on GCP. Preserved the product customizer's backend logic, ontology system, and WooCommerce integration throughout the migration.

Docker
Google Cloud
WooCommerce
MySQL
API
view case study →
010.

SAPHI - AI Agent Platform

Set up Google Cloud and OpenAI infrastructure for an AI Agent platform (SAPB1). Configured cloud-based inference pipelines, integrated OpenAI APIs with custom agent orchestration, and deployed the system on GCP for production use.

Google Cloud
OpenAI
Python
API
view case study →
011.

Multi-Tenant Deployment System

Redesigned a deployment system to add monitoring, automated rollback, and multi-tenant isolation. Built observability dashboards, implemented blue-green deployment strategies, and created tenant-scoped resource management on AWS.

AWS
Docker
Ansible
CDK
view case study →
012.

LLM Email Interface

Built an email-based interface for LLM interaction using AWS SES for email receiving, OpenAI Assistants API for conversation management, DynamoDB for state, and S3 for attachment storage. Implemented CloudWatch logging, CloudConvert for file processing, thread management for multi-turn conversations, and an AWS Amplify frontend with social login and custom prompt templates. 572 hours of development, $34,340 project value.

LLM
OpenAI
Python
DynamoDB
AWS S3
AWS Amplify
Lambda
view case study →
013.

CodelessGPT - AI Workflow Platform

Built a no-code AI workflow automation platform (codelessgpt.com). Users upload documents into a Library and Datasource system, compose reusable Prompt Templates, and wire together Actions (Gmail, API calls, web crawling, search) into automated Flows. The platform runs GPT-powered analysis pipelines over uploaded knowledge bases — e.g. resume screening against job descriptions, portfolio analysis, blood report analysis, pentest report review. Essentially a RAG-based workflow orchestration engine where non-technical users build and run AI-powered document processing pipelines without writing code.

OpenAI
React
Node.js
Python
Embeddings
API
view case study →
014.

AI-Native Hiring Funnel

The screening pipeline you're using right now. A single profile.json is the source of truth for this Gatsby site, an MCP server (8 tools incl. a semantic match_role engine), Schema.org + llms.txt + .well-known discovery, and a live serverless link-minter: paste a JD into a password-protected admin UI and get a short s.nishanttiwari.com link that returns a self-contained screening prompt — content-negotiated to serve a copy-and-paste page to browsers and raw text to AIs. No LLM in the request path, so it runs at ~$0. Surface propagation is regression-tracked in a custom agentic QA system (AQA).

Lambda
DynamoDB
Node.js
React
API
view case study →
more projects

what people say

Nishant is an amazing engineer! Even better than expected, both in quality and in time!

Upwork Client — Client — Secure Fleet Access

Thinking outside the box, agile, and just absolutely the best freelancer. I wish I could hire him fulltime!!!

Upwork Client — Client — Enterprise VPN

Self-motivated, determined and able to be left alone for long periods without supervision. Quick to learn and willing to turn his hand to anything.

UK Reference — Manager

Team is capable of building any feature and is technically sound. I have faith in the team and project leader on their capability to deliver.

Client Reference — Client — Aim Global

Very diligent with a solid understanding of user requirements and the technical ability to carry out changes using his skills and initiative. A pleasant and hardworking colleague.

Institute of Development Studies — Colleague — Sussex University

Excellent job. Very high level of quality. Extremely satisfied.

Client Reference — Client

Latest Articles

What is l33tPWN

Penetration Testing in the cloud via any web browser.| l33tpwn is a Zero configuration ethical hacking lab that runs in your browser. You will enjoy lag free full desktop experience of an attack machine. You will exploit multiple target web applications & operating systems.

react

October 3rd, 2020

more articles

Copyright © 2026 Nishant Tiwari All Rights Reserved