Email

Mike Elliott

Founder and CEO — AI systems for regulated industries. Marine Corps, then twenty-eight years in capital markets and regulated communication; the last four building AI, with one exit. I make new capability usable inside hard constraints.

Profile

I have spent my career in one place: inside hard constraints, making new capability usable without breaking them. It started as a Marine Corps ammunition technician, where procedure is why people go home. Then as a licensed Financial Advisor, where the rules were the job. Then seventeen years advising public companies on investor relations, when social media arrived and issuers needed channels that did not breach disclosure obligations. Now with AI.

The problem has not changed shape. A powerful new channel appears, the organisations that would benefit most are the ones most exposed if it goes wrong, and almost nobody understands both the technology and the constraint.

What I actually do is the whole loop, and it is three jobs most people split across three hires. I generate the demand — seventeen years of paid, organic and social campaigns before AI, and now AI-visibility and search work that puts a business in front of buyers who are asking a machine instead of Google. I build the system that answers them — Python and TypeScript, my own hands on the keyboard, from the voice agent to the database schema. And I govern it, so it survives contact with a regulator. Take any leg away and the other two stop being worth much: traffic you cannot answer, a system nobody finds, or a fast answer that gets the client sanctioned.

Today that means an agentic AI factory — custom agents and AI implementations for any vertical, sold hardest into law firms because a regulated buyer is where the governance work is worth most — and AI investor relations and governance for micro-cap issuers.

I work end to end and I write the code myself: the product, the agent orchestration layer, the governance standards clients are onboarded against, and the sales motion that puts them in front of buyers. What I am unusual at is knowing what an agent may claim, what must be logged, what a vendor record has to contain, and how to price it so the unit economics survive contact with real usage.

Experience

Jan 2026 – Founder & CEO

AlphaForge USA

Agentic AI factory · custom agents across verticals · alphaforgeusa.com

An agentic AI factory. We build custom AI agents and full AI implementations for any business vertical — voice intake, chat, outbound, research, back-office automation and the operations centers that hold them. Shipped into law, hospitality, land and real estate, and public markets, with a template library covering dental, HVAC and professional services.

Go-to-market is deliberately narrow while the platform is not. The named beachhead is Florida probate and estate-planning firms of 2–15 attorneys, because a regulated vertical with a specific compliance surface converts and retains far better than horizontal SMB. The agents underneath are general; the sales motion is not. That is a positioning choice, not a capability limit.

  • Ship and operate client "ops centers": production Next.js applications combining AI voice intake, chat, case or job flow and pipeline reporting, deployed per client and paid for monthly.
  • Authored the governance layer clients are onboarded against — an AI compliance standard, a legal-chatbot conduct standard, a security and safety control set, and a usage/cost governance regime.
  • Built a single-source-of-truth architecture so brand canon, pricing and ICP live in one governed repository that every downstream agent and application reads, rather than drifting across codebases.
Live client systems in production
2023 – 2024 Founder

AI Profit Lab

Acquired by Appsoft Technologies, 2024

Founded and built an educational platform teaching operators how to build and run AI agencies — the practical skill set for deploying AI agents into small and mid-sized businesses, from positioning and offer design through implementation and delivery.

  • Built the curriculum and platform for a market that was, at the time, long on hype and short on anyone who had actually shipped.
  • Sold to Appsoft Technologies in 2024 for an undisclosed sum, roughly eighteen months from founding.
  • The teaching work became the operating thesis for what followed: the agency skills I was documenting are the ones AlphaForge now sells as a service rather than a course.
Exit — acquired 2024
2022 – present Founder & CEO

InvestorConnect AI

AI investor relations & governance · micro-cap issuers · investorconnectai.com

Founded and still operating. An investor-relations and compliance platform serving public companies, including the Agent Fortress product — AI governance and control tooling for organizations that must answer to regulators and shareholders for how AI is used inside the business.

  • Productized compliance and governance for a buyer with genuine regulatory exposure, where an unsupported claim is a liability rather than a marketing problem.
  • Built the IR delivery pipeline: SEC EDGAR ingestion, issuer and CEO prospecting, qualification against financial-data APIs, and sprint-based delivery.
  • Runs as a separate brand with its own research and publishing surface, operating alongside AlphaForge rather than folded into it.
2005 – 2022 Founder & Principal

CEOLive

Investor relations consulting · public companies

Left Merrill Lynch to build my own investor-relations consultancy and ran it for seventeen years — sold it, delivered it and built it, with no one else to hand any part of it to. I interviewed public-company CEOs and advised issuers on how to communicate with the market — through the 2008 crisis, the arrival of social media, and the shift of investor communication from press release to platform.

  • Interviewed public-company CEOs at length, building a long-running body of issuer conversations and, with it, a working understanding of how the people who buy this kind of work actually think.
  • Social media governance for public companies — compliance-aligned deployment of issuer channels at a time when a single post could raise selective-disclosure questions. This is the direct precedent for the AI governance work I do now: same problem, different technology.
  • Ran the demand side for seventeen years. Every website, every organic social programme and every paid digital campaign across the client base was mine — strategy, build and spend. For issuers, where the audience is investors and every claim is disclosure-sensitive, which is a harder constraint to run growth inside than consumer marketing.
  • Built and ran the technology myself. The hands-on engineering habit AlphaForge runs on was formed here, not recently — I have been shipping my own web properties since long before it was a differentiator.
  • Owner-operator throughout: sales, delivery, technology and client relationships, with no one else to hand any of it to.
1998 – 2004 Financial Advisor

Financial Advisor — UBS Wealth Management · Merrill Lynch

FINRA Series 7, 52 & 63 — held during tenure, since lapsed

Six years as a licensed Financial Advisor across two global institutions, holding Series 7, 52 and 63 while registered through these firms. Registrations lapsed after I left the industry in 2004 and are not currently active.

Client suitability, disclosure discipline and supervised communication — the direct antecedent of the claim-language controls I now write for AI agents in regulated practice.

1988 – 1994 Staff Sergeant

United States Marine Corps

MOS 2311, Ammunition Technician · Operations Desert Shield and Desert Storm

Six years active duty to Staff Sergeant (E-6) — accelerated by three meritorious promotions, with two meritorious masts for outstanding performance. Ammunition accountability under combat conditions: custody, inspection and issue of ordnance, where procedure is not paperwork. Completed my degree while serving.

This is where the rest of the page starts. Everything above it is the same instinct — operate precisely inside constraints that exist for a reason — applied to securities, then disclosure, then AI.

Selected work

Systems I designed and built, running in production today. Each one is deployed, paid for, and carrying live traffic.

The AlphaForge Ops Center dashboard

The Ops Center

Next.js · Neon Postgres · Retell · deployed per client

The product clients pay for monthly, and it runs the whole lifecycle: a call or chat arrives, an agent qualifies it, it becomes a lead, then an intake form, then a matter under management — tracked to close and beyond, in one system rather than five.

The part I have not seen elsewhere is the Agents tab. Every agent on the account is instrumented: what it is doing right now, how much it has cost, and what it returned. It reconciles against the firm's SEO and paid-ad analytics, so the whole funnel is attributable — ad spend to phone call to signed matter.

Why it matters: almost every AI vendor sells a capability and leaves the client guessing whether it paid for itself. This answers that question per agent, in the client's own dashboard. When renewal comes up, the argument is already on the screen.

Watch it: alphaforgeusa.com/mike_elliott
Intake — live Marbury & Finch, PLLC
Caller Ringing

    The AI intake agent

    Retell · Twilio · ElevenLabs · answering live now

    This is a scripted illustration of the real call flow, looping beside you: the agent answers, listens, pulls the matter type and jurisdiction, runs a conflict check, and routes to an attorney. Firm and caller are fictional.

    Why it matters: watch what it refuses to do. It says it is an AI in its first sentence. It declines to give legal advice. It routes to a consultation rather than booking over an attorney's calendar. Those three constraints are the difference between a demo and something a regulated practice can actually put on its main line — and I wrote the conduct standard behind them.

    Call the real one — (407) 904‑9322
    Attribution — one matter ad spend → signed
    1. Search / AI answerA buyer asks Google — or asks an AI — and the firm is in the answer
    2. Paid or organic clickSource, medium and campaign captured on arrival
    3. Call or chatAgent answers, qualifies, extracts the matter
    4. Lead → intake → matterSame record, carried forward — not re-keyed
    5. Signed — and traceableCost and return resolve back to the campaign that started it
    The chain, not a dashboard screenshot. Each hop is a real field in the client’s own database.

    The demand side

    GEO & search visibility · paid campaigns · closed-loop attribution

    Answering the phone is worthless if nobody calls. Half of what I build is demand: search and AI-visibility work — auditing what ChatGPT, Perplexity and Google’s AI answers actually say about a business when a buyer asks, then fixing it — plus the paid campaigns and outbound that fill the top of the funnel.

    Why it matters: almost nobody closes the loop. Agencies report clicks; intake vendors report calls; neither can tell a principal which ad produced a signed matter. Because I own both ends, cost and return resolve against each other in the client’s own dashboard. That is the difference between a marketing invoice and a business case.

    Published AI-visibility audits for law, hospitality and property clients are live on the site.

    Also in production

    SystemDomainWhat it does
    Law firm ops center Legal — probate AI voice intake, chat, case flow and reporting for a practising law firm. Paid monthly; carries live call and chat volume.
    Law firm ops center Legal — estate Second firm deployment on the same platform, separately configured and monitored.
    Internal command center Sales & delivery CRM, pipeline, lead generation, proposal and delivery tooling operated with a business partner.
    IR sprint platform Public markets Issuer prospecting, EDGAR ingestion, qualification and delivery for investor-relations engagements.
    Hospitality & land ops Beverage · land Operator agents for a beverage-services company and a land/real-estate venture — lead response, booking and back-office flow.

    Technical skills

    AI & agent engineering

    Claude API and Claude Code, OpenAI, agent orchestration and multi-agent coordination, tool/function calling, structured output contracts, prompt and context engineering, retrieval-augmented generation, evaluation harnesses, local inference via Ollama.

    Voice & conversational AI

    Retell, Twilio, ElevenLabs. Production phone intake for regulated businesses: call handling, transcription, escalation paths, and the conduct standards governing what an AI agent may say to a member of the public.

    Application development

    TypeScript, React, Next.js App Router, Tailwind CSS, server components and actions. Vercel and Netlify deployment, preview/production promotion, environment and secret management.

    Backend & data

    Python (FastAPI, Pydantic, psycopg), Node. Neon serverless Postgres, SQL schema and migration design, Pinecone vector search, Vercel Blob, Stripe billing integration.

    Automation & infrastructure

    Scheduled agent fleets on launchd/cron, event-bus coordination between agents, health monitoring and failure reporting, atomic state writes, deterministic collectors designed to run at zero inference cost.

    AI governance & compliance

    Authored control sets for AI in regulated practice: approved and forbidden claim language, vendor records, retention policy, onboarding gates, and usage/cost governance. Built drift detection that fails a build when copy contradicts canon.

    Marketing & growth

    Seventeen years building and running client-facing web properties, organic social and paid digital campaigns — for issuers where every claim was disclosure-sensitive. Positioning, offer design, SEO and generative-engine visibility, lead generation and outbound.

    Integrations

    Slack, Discord, Gmail and Google Workspace, Perplexity and financial-data APIs, SEC EDGAR ingestion, webhook routing, MCP servers, browser automation.

    Operating discipline

    Cost-controlled AI architecture — separating deterministic collection from paid inference. Model routing by task class and reversibility. Single source of truth for pricing, positioning and ICP across every repository.

    Domain expertise

    Legal

    Florida probate and estate-planning practice operations — firms of 2–15 attorneys. Intake, matter flow, and the conduct rules governing AI contact with prospective clients, including unauthorized-practice and attorney-advertising constraints that determine what an automated system may say. Based in Orlando and selling into the Florida bar, so the buyers are local and reachable in person.

    Public markets

    Twenty-four years of continuous involvement — advisor, then IR consultant, now vendor. Issuer communications and disclosure-sensitive content, social media governance under selective-disclosure constraints, EDGAR data, and AI governance for companies accountable to regulators and shareholders. I have sat on the advisor side, the consultant side and the technology side of the same table.

    Capital markets

    Economics and finance training, then six years in wealth management at UBS and Merrill Lynch. Client suitability, disclosure discipline, and the habit of treating a claim as something that has to survive review — the same reflex applied to what an AI system is permitted to assert.

    Selected written work

    Governance and standards documents authored for client-facing AI deployment. Available on request; useful evidence for buyers who need to see that the compliance layer exists.

    DocumentCovers
    AI Compliance Standard Control set for law-firm AI agents: rules built to, open-item pattern, onboarding gate, approved and forbidden claim language, vendor record, retention.
    Legal Chatbot Standard Conduct standard for AI in direct contact with prospective legal clients, with an automated tone-lint rule set.
    Security & Safety Security posture and safety controls for law-firm ops centers.
    Usage & Cost Governance Usage limits and cost governance so AI spend stays predictable at the client level.

    Education

    1994

    Iowa State University

    B.A. Economics · minor in Finance · completed while serving on active duty