Skip to résumé

Sheet 01 · Chad Koslovsky

Full-Stack Product Engineer

Senior Engineer: Agent Academy · Focused

I build TypeScript products with clients, from the first conversation through the interface, API and release. I bring that same hands-on approach to AI-assisted workflows: give tools clear boundaries, review what they do, and keep the result understandable to the team that will own it.

Sheet 02

Cover letter

I’m a self-taught engineer, and the problems I’ve learned the most from are the ones I had to work through with the people who would live with the result. That’s what caught my attention about Agent Academy.

Read full cover letterClose cover letter

Hi Focused team,

I’m a self-taught engineer, and the problems I’ve learned the most from are the ones I had to work through with the people who would live with the result. That’s what caught my attention about Agent Academy. You’re not just training people to build agents. You’re asking them to pair, figure out what good behavior looks like, and help a client team own what they build together.

At Technomancy IT, I take client software from an initial conversation through design, implementation and support. One long-term engagement involved a contactless radar device. I handled full-stack engineering and DevOps setup and support alongside the data science and ML engineers. My contribution wasn’t the model or the device. It was making the software and environment around their work usable and supportable. At Front Range Internet, I learned the other side of that handoff running the NOC. The team handled first-line issues, and I took the escalations they couldn’t resolve. Making procedures clear and verifying recovery were part of keeping that work from depending on what one engineer happened to know.

I’ve been applying that thinking to AI-assisted development and tool integrations. I built access-controlled connectors for workflows involving Slack, Google Workspace and Atlassian, and I use coding agents in my own work. One thing I keep coming back to is that an agent can sound certain when it doesn’t have enough information. I want to get better at evaluating when it should take the next step, when it should ask a better question, and when the answer should be an ordinary deterministic workflow instead. I’m especially interested in making those decisions visible enough that a client engineer can challenge the behavior and improve it with me.

I’d bring you experience shipping TypeScript software, staying close to the people using it, and taking responsibility for what happens after release. I’d like to build on that through the Academy’s evaluation and context engineering work, then put it to use alongside your teams and clients.

Thanks, Chad Koslovsky

Sheet 03

Experience

Technomancy IT LLC · 2015-Present

Founder / Product Engineer / Systems Lead

Build TypeScript products with clients, from discovery and interface design to typed APIs, release and support. Built access-controlled MCP and SaaS integrations; handled full-stack work and DevOps support alongside a client data science team.

Deep dive

I work directly with clients to turn an operational problem into software, then stay involved through release and support. I define the interface and data model, build the application and put the operating pieces around it so the next change doesn’t require starting over. On a long-term radar sensing engagement, I handled full-stack engineering and set up and supported DevOps for a data science and AI team; my contribution was the software and delivery environment, not the sensing model or device firmware.

For AI and SaaS integrations, I use authenticated service boundaries and access-controlled connectors. I built integrations spanning Slack, Google Workspace and Atlassian as well as internal tools. Keeping the tool’s permitted actions explicit makes it easier to review a workflow before it can touch a client system.

Front Range Internet Inc. · 2015-2020

Senior Systems Engineer / Datacenter Architect

Ran the NOC, took engineering escalations and documented operating and recovery procedures the team could use. Supported web services and migrated workloads from Hyper-V and VMware ESXi into Docker and Kubernetes.

Deep dive

I ran the network operations center and took escalations its first-line team could not resolve. That included making monitoring, recovery verification and operating procedures understandable to the people responding to incidents. Later in my tenure, I carried sole systems-engineering responsibility for about a year while the NOC continued first-line coverage.

I moved workloads from Hyper-V and VMware ESXi into Docker and Kubernetes and supported web services. I also ran SOC 1 and SOC 2 audits; part of the work was turning technical controls into evidence that someone outside the engineering team could follow.

Office Evolution · 2014-2015

Systems / Software Engineer

Built SQL- and MongoDB-backed applications connecting CRM, provisioning, onboarding and service operations. Worked across departments to make handoffs explicit.

Deep dive

I built SQL- and MongoDB-backed applications across CRM, provisioning, onboarding and service operations. I worked across departments and with executive stakeholders to represent the handoffs in shared software rather than make each office invent its own process. The job joined writing the application with making sure its surrounding systems could support how the teams used it.

PlumChoice / Everon · 2012-2014

Network & Systems Engineer

Built diagnostic tools for recurring support issues and worked directly with managed-service clients on networks, systems and escalations.

Deep dive

I worked on networks and systems across managed-service clients, taking on-call issues and working directly with users and business owners. Alongside that support I built diagnostic tooling for recurring issues. That work taught me to explain what a diagnostic result meant for the business, not just which device had failed.

IBM · 2010-2012

Infrastructure and Systems Focal

Coordinated application and infrastructure changes across teams, automated audit and backup workflows and managed shared virtualization test labs.

Deep dive

I was the infrastructure and systems focal across application, infrastructure and vendor teams. I automated audit and backup workflows and managed virtualization test labs and hosted desktop platforms. Keeping change and incident work coordinated mattered as much as getting an individual script or environment working.

Sheet 04

Selected work

Completed client contribution

ProofMark · Document signing

Designed a signing workflow across identity, routing, PDF fields and audit history, with React, tRPC and PostgreSQL. Kept verification and fallback states explicit.

Deep dive

I designed the signing workflow as a completed client engineering contribution. It connected signing identity, routing, document fields, verification and audit history, using Next.js, React, tRPC, Drizzle and PostgreSQL alongside Rust work. Explicit fallback paths made verification state part of the application rather than a thing a reader had to infer.

ProofMark product site ↗

Independent project · Early development

Kerf · TSX framework for people and coding agents

Built a compiler-backed UI framework with fine-grained DOM updates. Diagnostics explain mistakes a developer or agent can make when TSX looks familiar but the reactivity works differently.

Deep dive

Kerf is an early-stage TypeScript and TSX framework that includes a compiler and the Warp UI toolkit. A component runs as a setup function rather than re-running the whole component for every update; the compiler and runtime keep track of which UI bindings need to react. I wanted mistakes to give a developer or a coding agent something specific to fix, so the checks identify patterns such as a value read once when the UI is meant to keep updating.

Kerf project preview ↗

Founder-built product

Technomancy · Client support platform

Built a Next.js and PostgreSQL operating layer for tickets, CRM, scheduling and billing. Tenant-aware services distinguish client operations from internal work.

Deep dive

This platform joins client support, ticketing, scheduling, customer records and billing in a TypeScript and PostgreSQL application. Shared services let these workflows exchange the data they need without making client and internal access interchangeable. The documented release script uses Git-derived image tags, database migrations and post-deploy health checks, making it possible to trace a rollout to its image and follow up when a check fails.

Platform overview ↗

Sheet 05

Capabilities

  • TypeScript · React · Next.js
  • API design · tRPC · PostgreSQL
  • MCP integrations · AI-assisted development
  • Docker · CI/CD · release verification
  • Client discovery · cross-functional delivery