You are not selling AI theater internally.
Your visible LinkedIn activity pushes AI across the full SDLC, from story writing to testing and deployment, with governance built into the workflow.
src / S2 / LinkedIn activity inspectionVincent, the spare-wire post made this feel very real: you have lived through enough infrastructure eras to know when a shiny AI layer is just another bin of old cable.
Vincent Miceli / VP, AI and Automation / Point C
Your posts keep coming back to the same point: coding acceleration is not enough if requirements, reviews, tests, deployment, and governance still run the old way. Point C has the right kind of operating surface for that thesis.
Two things about us
We are a small AI automation services company with a US-based delivery team. We like building relationships in person, and we will fly to you when that helps get the workflow set up right.
Our founders are ex-Google, Meta, and Snowflake engineers who spent over 10 years building AI and data products used by millions. We start with a 14-day free engagement. If we do a good job, all we ask is a good word.
What we noticed before the call
Three signals from public research or explicit hypotheses. If any of this is off, correct it on the call.
Your visible LinkedIn activity pushes AI across the full SDLC, from story writing to testing and deployment, with governance built into the workflow.
src / S2 / LinkedIn activity inspectionThe public site points to brokers, employers, members, providers, cost management, reporting, compliance navigation, and plan administration.
src / S4 / Point C public websiteCareers showed work across claims, system configuration and plan build, stop-loss sales support, account management, finance, and EDI or IT.
src / S5 / Point C Greenhouse jobsThree workflow opportunities
Each workflow is scoped as a managed service with deterministic gates, evidence logs, and human review where judgement matters.
maps to OpenNash Executive AI Stack
| Before | AI writes code faster, but requirements, reviews, tests, deployment notes, and evidence packs still move by hand. |
| Agent role | Turn stories into acceptance tests, generate review evidence, prepare deployment checklists, and track what changed. |
| Human review | Engineering or product owners review exceptions, risk flags, and final release gates. |
| Measured | Cycle time, review rework, test coverage, escaped defects, and release evidence completeness. |
maps to OpenNash Workflow Automation
| Before | Employer, broker, and internal changes arrive as documents, emails, portal requests, and rules that need translation into build tasks. |
| Agent role | Extract plan rules, create a config checklist, draft test cases, compare against prior builds, and flag ambiguous clauses. |
| Human review | Plan-build owners review ambiguous rules, compliance-sensitive language, and final configuration approval. |
| Measured | Build cycle time, config defects, manual touch count, exception rate, and rework avoided. |
maps to OpenNash Voice & Messaging
| Before | Portal issues, claim questions, eligibility questions, EDI exceptions, and stop-loss support requests compete for the same human attention. |
| Agent role | Classify the request, pull the right records, draft the next action, route regulated cases, and log every step. |
| Human review | Humans handle protected decisions, escalations, and anything outside the approved playbook. |
| Measured | Time to first useful response, backlog age, clean resolution rate, escalation rate, and audit completeness. |
What ships
One Point C workflow decomposed into inputs, systems, decisions, risks, and exception gates.
Rules for PII handling, approval gates, test evidence, and when humans must step in.
A reviewed workflow that drafts, checks, routes, and logs real operational work.
Every action recorded so engineering, ops, and compliance can inspect what happened.
Cycle time, exception rate, rework, backlog, and value captured in plain numbers.
How Point C owns, changes, and retires the workflow after the pilot.
Pilot terms
You already understand the hard part: AI does not fix the system unless it becomes part of the system.
OpenNash starts with one workflow, not a platform rollout. We define the acceptance tests, wire the workflow with review gates, and keep the operational evidence visible.
If the workflow does not pass the tests, you do not pay. If it does, it becomes managed work with logs, metrics, and Point C ownership.
Thirty minutes. Confirm the operating picture, pick one workflow, and decide what evidence would make it worth keeping.