A live MCP server + A2A agent for AI assistants
Teach your AI about Jake Gaylor.
Point your assistant at this server and it can read my resume, score my fit against your job description, build the interview, book a call on my real calendar, and email me — no copy-pasting, no stale PDFs.
https://ai.jakegaylor.com/mcp
What your assistant can do here
Connect once, then ask in plain English. The server exposes tools for the whole evaluation — from first look to first contact.
Assess my fit
Paste your job description and get an honest read on where I'm strong and where I'm not, scored against my real experience.
“Is Jake a good fit for this role? [paste job description]”
Build the interview
Generate phone screens, technical deep-dives, system design sessions, or behavioral questions tailored to my background and your focus areas.
“Create a phone screen for Jake focused on Kubernetes and API design.”
Pull the source material
My resume, GitHub, LinkedIn, and website content, served as structured resources your assistant reads directly — always current.
“Summarize Jake's career highlights.”
Skip the screening call
Location, remote preference, work authorization, comp range, and target roles as structured JSON — the whole logistics screen, answered before anyone talks.
“What are Jake's screening preferences?”
Book a call
Live availability from my real calendar, booked in one exchange. Every booking is pending until I confirm it — you get the outcome by email.
“When is Jake free this week? Book Thursday at 11.”
Reach out
Looks like a match? Your assistant can email me directly from the conversation — you review it, the server sends it.
“Draft and send Jake an email about our staff platform engineer role.”
Connect in under a minute
The server is already hosted and running — nothing to install. Add the endpoint to any MCP-capable client, use the npm package for clients that only speak stdio, or let an A2A-capable agent discover the agent card on its own.
Streamable HTTP
Recommendedhttps://ai.jakegaylor.com/mcp
SSE
Legacy clientshttps://ai.jakegaylor.com/sse
A2A Agent Card
Agent-to-agenthttps://ai.jakegaylor.com/.well-known/agent-card.json
Client setup
Pick your client. Clients without remote MCP support can run the same server locally via the @jhgaylor/me-mcp npm package.
No MCP client? Copy this instead
Paste the text below into any AI assistant to give it my full background — no tools or special connections needed. Also available at ai.jakegaylor.com/llms.txt.
# Jake Gaylor ## Senior / Staff Engineer & Engineering Leader | I build systems a stranger can run on day one [email protected] | 720.453.3994 | https://jakegaylor.com | https://github.com/jhgaylor | https://www.linkedin.com/in/jhgaylor/ Boston, MA · will relocate ## Summary I build systems a stranger can run on day one. 15+ years across SaaS, AI, infrastructure, and IoT. At Cloaked I led the move off a legacy PaaS onto AWS EKS, made deploys 30x faster in six months, and ran that platform as its one-person ops team while the product grew from hundreds of users to hundreds of thousands under SOC 2 and ISO 27001. At CyberGRX I rebuilt a quarterly, manual deploy process and had every engineer shipping on it within 3 days. Agents have been in my production systems since LangChain was new, and today a fleet of coding agents builds Managoat, the hosted agent platform I run, under my review. ## Professional Experience ### Ravi | Co-Founder | Jan 2026 - May 2026 https://ravi.id Cofounded an agent-native SaaS where AI agents manage their own accounts, secrets, and 2FA. - Raised a $400K pre-seed from South Park Commons on the usage data from a v1 instrumented from day one - Grew to hundreds of users - Shipped it end to end, a web SaaS and a Flutter app on the App Store, from prototypes started in 2024 - Designed agent-native account, secret, and 2FA management so AI agents can operate their own identities ### Block Party (acquired by DeleteMe) | Staff Software Engineer | May 2025 - Jan 2026 https://www.blockpartyapp.com/ Built internal AI and developer tooling for the engineering org until Block Party was acquired by DeleteMe in January 2026. - Designed and built the framework behind the company's internal AI coding bots, treating the agent as another consumer of the developer platform ### Cloaked Inc | Staff Platform Engineer → Staff Product Engineer | Sep 2022 - May 2025 https://cloaked.app Started in platform engineering with the AWS EKS migration, GitOps CI/CD, and compliance, then moved into product to build autonomous AI agents and the ideation portal behind a sticky, AI-powered onboarding experience. - Led the migration from a legacy PaaS to AWS EKS with a multi-account model meeting SOC 2, ISO 27001, and ISO 27701 requirements, and it came out cheaper with more headroom than what it replaced - Built a GitOps-based CI/CD pipeline that accelerated deployments 30x in my first 6 months - Carried that platform from hundreds of users to hundreds of thousands without a rebuild, as its solo ops team - Built an end-to-end product-analytics stack that enabled data-driven decisions across the company - Built autonomous AI agents with LangChain that log in to user accounts to change passwords and swap the email on file, then made them a core fundraising narrative and an onboarding experience that lifted retention - Launched a cross-functional ideation portal nearly the whole org used, one board from idea to usage report for design, product, engineering, and executives ### The Onward Store Steakhouse | Owner / Co-GM | Oct 2021 - Jan 2024 Owned and ran a full-service steakhouse with my wife, 7 staff and $500K a year in business revenue, P&L on me. - Ran the full business lifecycle, from spotting the opportunity through acquisition to a $500K/yr operation - Hired, trained, and scheduled a team of 7, learning what each person needed instead of running one playbook over all of them - Owned the P&L, budgets, and financial reporting, holding margin through inventory management and cost control - Built the operating procedures behind maintenance, vendors, and quality assurance that kept service consistent shift to shift - Ran marketing and community engagement that grew local awareness and repeat patronage ### Inception Health | Staff SWE in Developer Experience | Jun 2020 - Jun 2022 Designed the AWS CDK patterns that product and mobile teams used to ship HIPAA-compliant patient-care microservices themselves, and left both teams running their own deploys. - Led delivery of the AWS CDK infrastructure behind a HIPAA-compliant patient care microservices platform - Designed IaC patterns application teams consumed directly, so a team could ship a compliant service without a platform engineer in the loop - Built mobile CI/CD that turned TestFlight and App Store releases into self-serve deploys from git - Collaborated with AWS and partner teams to integrate managed services and accelerate platform development - Trained and mentored the product and mobile teams on CDK, serverless, and CI/CD until they ran the platform without me ### CyberGRX (now ProcessUnity) | Senior Platform Engineer | Dec 2018 - Jun 2020 https://processunity.com/ Rebuilt build and deploy at a company where shipping was a quarterly manual event, and had roughly 50 engineers across several teams deploying on demand within 3 days. - Wrote a Go operator on the CoreOS operator SDK that ran blue/green rollouts as a custom resource, keeping the previous version a selector flip away so a bad deploy was undone in under a minute - Redesigned the build and deploy process, presented it to the org, and had every engineer shipping with it inside 3 days - Led the vision for cloud-native CI/CD pipelines on AWS and Kubernetes - Mentored software engineers in best practices for building and deploying software - Grew into a Staff Engineer role by owning large-scope initiatives, collaborating with the Tech Lead to drive interactions across the organization and unblock progress ### Cardfree | Senior Cloud Platform Engineer | Jan 2018 - Dec 2018 Ran a large-scale PCI-compliant C# installation in AWS and took the hand-built hosts out of it, with every Windows and Linux server coming from one Packer build. - Architected a hybrid AMI build and deploy platform with Packer and Vagrant, so a drifted host behind the C# and Ruby services got replaced from the image rather than repaired - Standardized infrastructure-as-code practices to pave the way for the open-source .NET migration - Replaced ad-hoc Slack firefighting with a structured PCI-compliant incident response program, structured enough for the auditors and light enough that engineers actually followed it ### Protectwise, Inc (Acquired by Verizon) | Senior DevOps Engineer | Aug 2016 - Jan 2018 SRE for a petabyte-scale ingestion platform, thousands of Cassandra nodes and Kafka pipelines on a $10M+/yr AWS environment run entirely as code. - Managed SRE for a massive-scale data ingestion platform with thousands of Cassandra nodes and petabytes in S3 - Administered a $10M+/yr production AWS environment with Chef and infrastructure as code - Held the platform steady under a product that shipped dozens of times a day and could not miss data ### Food Service Warehouse | DevOps Engineer | Jun 2015 - Mar 2016 Ran production Kubernetes on vSphere in 2015, the year it hit 1.0, and put the Node.js services on the fastest release cycle in the company. - Implemented Kubernetes clusters on vSphere using CoreOS, fleet, etcd, and flannel - Brought Docker and Kubernetes to a C# and vSphere shop and made containers the release path for the first Node.js backend teams ### Mississippi State University | Software Developer | Jun 2010 - Jun 2014 Created grant proposals and custom software for research initiatives - Created grant proposals to fund custom software for university research initiatives - Developed software for tracking 'human sensors' through natural disasters including Hurricane Sandy ## Skills ### AI & Agent Engineering MCP Servers & SDKs, Agent Orchestration & Sandboxing, LLM Integration, Agent-Native Product Design, Agent Memory, LangChain ### Cloud Infrastructure AWS, Kubernetes (EKS, GKE, k3s), Terraform, Tailscale, Cloudflare, Infrastructure as Code ### GitOps & CI/CD Flux CD, ArgoCD, Argo Workflows, Github Actions, SOPS, Continuous Deployment ### Programming Languages TypeScript, Javascript, Python, Go, Bash, Dart ### Data & Storage PostgreSQL, Redis, Cassandra, DynamoDB, S3-compatible Object Storage, MongoDB ### Distributed Systems Kafka, RabbitMQ, AWS IoT, Event-Driven Architecture ### Observability Prometheus, Grafana, Loki, Pagerduty, CloudWatch, Product Analytics ### Product Engineering Data-driven Iteration, Feature Development, User Feedback Loops, MVP Development, Product Analytics ### Product & Business Roadmapping, Customer Discovery, P&L Management, Go-to-Market Strategy, Fundraising ## Projects ### Managoat https://managoat.com A hosted platform for running fleets of sandboxed coding agents, sold by the agent hour and built by the agents it orchestrates; the server under it, Fountain, is open source - 400+ PRs merged in its first three months, all agent-written; I set the roadmap and review the work - Ships under BinaryBourbon, the GitHub identity my agent fleet works as - Metered billing at $0.25 an agent hour, a published concurrency ceiling that refuses rather than queues, and a year of audit events - Sandboxes boot holding one short-lived signed identity, so no long-lived secret ever enters the box and teardown invalidates everything it held - An inference proxy injects the model key after the request leaves the sandbox and bills every call to the verified identity, with a budget where the open-ended bill used to be - Every feature on three surfaces, a web UI for debugging, a REST API for CI/CD, and a Homebrew CLI with manifest-driven fountain apply - Answers ACP, AG-UI, and MCP in both directions, so any client already in use can drive the fleet ### Production-Grade Homelab https://jakegaylor.com/homelab/ A three-node k3s cluster on M4 Mac minis serving production traffic from home, GitOps-managed, fully observable, photo tour included - Flux CD reconciles every manifest from git; secrets live encrypted in-repo with SOPS + age - Tailscale WireGuard mesh between Lima VMs with direct LAN paths; Traefik, cert-manager, and oauth2-proxy at the edge - Longhorn replicated block storage, Garage S3, CloudNativePG, and Prometheus/Grafana/Loki with alerts pushed to my phone ### ai.jakegaylor.com https://ai.jakegaylor.com/ A live MCP server and A2A agent card that let a screening assistant interview me directly - Any MCP client can read my resume, score my fit against a job description, and build the interview - The assistant can draft an email to me from the conversation, the human reviews it, and the server sends it - Streamable HTTP and SSE transports, an npm package for stdio-only clients, and llms.txt for everyone else - Runs on the homelab k3s cluster, deployed by Flux from git like everything else there ### MCP Working Group Contributions https://github.com/jhgaylor Early participant in Model Context Protocol (MCP) community development - Was a member of the MCP Hosting Working Group - Developed strategies for deploying and running MCP servers at scale - Contributed to the Dart SDK and an Express tool for binding to MCP servers - Built MCP servers for Hirebase, candidate job search, and text extraction - Created a GitOps tool for MCP deployments - Still operate a public Echo MCP server for community testing ### CareerFlick LLM-Driven Mobile App Developed a mobile application with a Tinder-like LLM interface - Built intuitive swipe-based interface for interacting with LLMs - Implemented AI-driven content generation and response system - Created responsive mobile-first user experience ### Hirebase MCP Server https://github.com/jhgaylor/hirebase-mcp A Python MCP server for Hirebase.org's API - Built a Python-based Model Context Protocol (MCP) server for searching jobs on Hirebase.org - Dockerized the service and set up GitOps CI/CD pipelines for automated builds and releases ### Express MCP Handler https://github.com/jhgaylor/express-mcp-handler Package to simplify MCP handling in Express.js - Developed Express middleware in TypeScript to streamline MCP request handling - Authored full type definitions and robust error handling for production-grade reliability - Configured CI/CD pipelines to automate code validation and package publishing ### Node Candidate MCP Server https://github.com/jhgaylor/node-candidate-mcp-server A TypeScript library to build customizable candidate MCP servers - Authored a reusable library enabling developers to integrate their own candidate data into MCP flows - Published as an npm package for easy adoption and extension by other teams ### mcp_dart https://github.com/leehack/mcp_dart Dart SDK implementation of the Model Context Protocol (MCP) - Contributed stream based support for in-process MCP servers - Add support for running MCP servers in flutter apps - Authored Dart examples to accelerate SDK adoption ## References "Jake is a force for good unlike any other. He migrated our infra to a cheaper and more scalable system, crafted a CI pipeline that accelerated deploys by 30x, and taught the whole company how to make data-driven decisions by building and showcasing an entire product analytics stack. And that was just his first 6 months." - Aaron Biller, Director of Platform Engineering, Cloaked "Jake and I ran the AWS behind a petabyte-scale security platform. Thousands of Cassandra nodes, Kafka pipelines, a $10M-a-year bill, all of it managed with Chef and deliberately boring tech. I'd sign up to do it with him again tomorrow." - Alex Artigues, DevOps Engineer, Protectwise "I watched Jake redo a build and deploy process over the span of 3 days, whip up a presentation, and get every engineer in the org up to speed using it to deploy new services, a moment I'll never forget because our productivity EXPLODED afterwards." - Stephen Bapple, Software Engineer, CyberGRX "We were deploying C# on vSphere when Jake joined our first Node.js backend team. He had been using Docker, we identified Kubernetes as our next step, and Jake hit the ground running. Our Node services had the fastest release cycle in the company." - Architect, Food Service Warehouse, 2015 "Fountain is dope. Jake built one of the most GOATed models for everyone wanting self sovereignty over agent compute while getting the GrokBot like experience using Buzz." - Akeem Jenkins, Director of Product Engineering, Piedmont Global "Jake worked in the pit with [the operations team] and everyday came in with killer energy. Everyone else told us how to work around problems. Jake built us better tools." - Magician, Magic "I also wanted to second (third? Fourth?) Jake's skills here. I don't have anything unique to add as everyone else has already covered it, but this man will help you achieve what you need." - Senior Platform Engineer, Inception Health
Under the hood
This site speaks two open agent protocols: the
Model Context Protocol
for AI assistants a human has configured, and
Agent2Agent (A2A)
for autonomous agents that discover the
agent card
on their own. The MCP server is open source:
@jhgaylor/candidate-mcp-server.
Tools — actions your assistant can take
assess_role_fit
generate_interview_questions
contact_candidate
get_resume_text / get_resume_url
get_github_url / get_linkedin_url / get_website_url
get_website_text
get_candidate_preferences
get_availability / book_intro_call
Resources & prompts — data it can read
candidate-info://resume-text
candidate-info://website-text
candidate-info://*-url
resume-url, linkedin-url, github-url, website-url).
Built-in prompts
evaluate_job_fit, generate_phone_screen, get_candidate_background,
summarize_career_highlights, assess_tech_proficiency,
assess_product_collaboration, and assess_startup_fit.
A2A skills — what an autonomous agent gets, no configuration required
about-jake
assess-role-fit
candidate-preferences
schedule-intro-call
BOOK: <slot> | <email> | <name> books a 30-minute intro call. Bookings are pending until I confirm; the booker gets the outcome by email. Humans can book at cal.jakegaylor.com.
contact-jake
CONTACT: are delivered to my inbox by email. Include a reply address.
connect-via-mcp
Let your assistant run the first screen.
Connect it, paste your job description, and find out in minutes whether we should talk. If the answer is yes — it knows how to reach me.