Discuss a problem →
Menu

Systems that know your business. And move it forward.

Palmetto Interactive helps executives turn fragmented knowledge and high-value workflows into trusted AI systems. We combine agentic AI, knowledge graphs, ontology engineering, and production architecture from strategy through delivery and handoff.

Pick a question below. The agent's path lights up in the graph, with its sources.

Ask it a question

01 · Client triggers

Start with the business constraint.
Then earn the technology.

Palmetto works with executive, technology, data, product, and operations leaders when a consequential decision or knowledge-intensive workflow needs a practical path forward.

  1. 01

    An AI pilot cannot cross the production gap.

    The demo works, but data quality, permissions, evaluation, economics, or operating ownership remain unresolved.

    Decision · scale, revise, buy, or stop
  2. 02

    Critical knowledge is fragmented.

    Policies, documents, data, and expert judgment live in different systems. Teams do not always mean the same thing by the same words.

    Measure · answer quality, provenance, rework
  3. 03

    A valuable workflow is still manual.

    People spend their time gathering context, reconciling sources, routing exceptions, and moving work between systems.

    Measure · cycle time, touches, exceptions
  4. 04

    Leadership needs an independent decision.

    The organization needs to know what to fund, what to integrate, what to govern, and what not to build before more budget is committed.

    Decision · build, buy, sequence, or defer
Executive teamsfund the right workTechnology & data leadersmake it operableProduct & operations leadersmake it useful

02 · Capabilities

Four connected disciplines.
One commercial outcome.

Agents need trustworthy knowledge. Knowledge graphs need a job. We connect both to the architecture and ownership required to ship.

  1. 01

    Agentic AI systems

    Agents that use real business tools to finish useful work, with permissions, evaluation, and human control.

    • workflow agents & copilots
    • tool use · orchestration · evals
    • human approval where it matters
    • observability · safe rollout
  2. 02

    Knowledge graphs & ontologies

    Shared definitions for data, policies, and expertise so people and AI reason over the same context.

    • ontology & semantic modeling
    • knowledge graph architecture
    • entity resolution · provenance
    • GraphRAG · conversational data
  3. 03

    Data & AI architecture

    Retrieval, integration, governance, and evaluation that still work after the demo.

    • AI-ready data foundations
    • retrieval · memory · provenance
    • integration & governance
    • platform & operating model
  4. 04

    AI products & workflow automation

    From a validated workflow to a product the team can run: pilot, integration, rollout, and handoff.

    • workflow & product framing
    • production-shaped pilots
    • integrations · interfaces · controls
    • rollout · runbooks · handoff

03 · Industries

Where the work is specific.

Three industries. Named workflows. The linked examples are anonymized experience, not endorsements.

Financial Services

When a critical service fails, the map of what else fails is still in spreadsheets.

  • Operational-resilience mapping

    Which services ride on which systems, teams, and vendors. Query it. Do not rebuild it when a regulator asks.

  • Operational-risk graphs

    Risks, controls, processes, and incidents in one structure. Not four taxonomies stitched together at reporting time.

  • KYC/AML entity resolution

    Customer, entity, and account relationships visible for review. Not reconstructed in the middle of an investigation.

Insurance

Underwriters and adjusters still rebuild the same policy file by hand.

  • Underwriting assist

    Submission, risk, and prior claims in one view so the underwriter can decide.

  • Claims triage

    First notice arrives with the context the adjuster needs to take the file.

  • Policy & coverage semantics

    One governed definition of coverage across lines of business. Analytics and AI use the same language.

Manufacturing

The part, the fault code, and the last fix still live in people and PDFs.

  • Field-service copilot

    Answers from equipment history and manuals, not a search box over a document pile.

  • Parts & BOM intelligence

    Which part fits which machine, across models and revisions, without plant-by-plant reconciliation.

  • Supply-chain resilience

    Which suppliers and lines fail together when a part goes short.

04 · Selected work

Knowledge systems delivered where consequences are real.

Three anonymized programs. Not proof, references, or endorsements.

claims-agent / run 0412
  1. Question from the adjuster

    Does the flood exclusion apply to claim 2026-0412?

  2. Walked the policy graph

    Claim to Policy HO-3 to Rider R-17 to Exclusion 4.2

  3. Found three sources

    HO-3 §4.2Flood and surface water are excluded.Rider R-17Adds sewer backup. Silent on flood.Claim note 3Water entered at grade during the storm surge.

  4. Drafted the coverage letter

    The loss is excluded under §4.2. Rider R-17 does not extend coverage to surface water. Citations attached.

  5. Waiting for a person

  1. 01 Financial Services

    Operational-resilience mapping

    What breaks if this service fails, queryable, not reconstructed.

    Global asset manager

  2. 02 Financial Services

    Operational-risk graphs

    Risk, control, process, and incident data in one assessable structure.

    North American bank

  3. 03 Manufacturing

    Agentic AI deployment on an enterprise graph

    Agents working from one governed map of plant and business data.

    Global agricultural equipment manufacturer

More work across continuity, operational risk, insurance semantics, and current Palmetto delivery.

Selected work →

05 · Engagements

Buy the decision and capability
the moment requires.

Start with a defined outcome and a focused path to it. Each engagement has clear deliverables, working sessions, decision points, and an ownership plan.

  1. 2–6 weeks

    Engagement 01

    Executive AI & knowledge diagnostic

    What should we fund, and what should we avoid?

    Buyer receives

    • opportunity and workflow map
    • value, readiness, and risk case
    • target architecture and sequenced roadmap
  2. 4–8 weeks

    Engagement 02

    Knowledge foundation sprint

    What must people and systems agree on?

    Buyer receives

    • competency questions and domain model
    • ontology, graph, and source mapping
    • provenance and stewardship plan
  3. 6–12 weeks

    Engagement 03

    Production-shaped agent pilot

    Can this workflow create measurable value safely?

    Buyer receives

    • working slice with representative data
    • evaluation and control baseline
    • cost, scale, revise, or stop recommendation
  4. 3–6 months

    Engagement 04

    Scale & operating partnership

    How do we productionize and transfer ownership?

    Buyer receives

    • integrations and production controls
    • observability, runbooks, and enablement
    • ownership transfer or retained leadership

06 · Palmetto product

Coming soon
Quillby Palmetto Interactive

You remember the thing.
Not where it lives.

A decision from an old project. A photo you can almost picture. A person you recognize but cannot place. Quill searches the files, photos, projects, and relationships behind your memory. Then it brings back the evidence.

Preview Quill →
The problem

You remember the context, not the location.

  1. 01Decisionburied in work
  2. 02Memorymissing a name
  3. 03Connectionmissing the why
Quill

Ask naturally. Get the trail back.

Matt Lucas with his family

Matt Lucas with his family in the Lowcountry

07 · Founder & principal

Executive perspective. Builder's accountability.

Matt Lucas is a Charleston-area technology executive and the founder of Palmetto Interactive. His experience includes Field CTO for Financial Services at Stardog, nearly ten years at Morgan Stanley including as an Executive Director of global technology delivery, and earlier architecture roles at IBM and OpenPages.

He stays accountable from the first framing conversation through architecture, delivery decisions, and handoff. Palmetto works with leadership teams, operators, internal technologists, and chosen vendors so the client can own what ships.

  • Senior leadership from framing to handoff
  • On-site collaboration across major U.S. markets
  • Built to strengthen the client's own team
Read Matt's full bio →Discuss a priority →LinkedIn ↗

08 · Insights

The thinking behind
the delivery.

Practical field notes from Palmetto, alongside selected talks and writing from Matt's enterprise technology career.

Palmetto field notes

Product previewYou Remember the Thing. Not Where It Lives. Meet Quill.Introducing Quill, a coming-soon membl product for finding half-remembered decisions, moments, people, and projects, with the evidence attached.HiringHow to Hire an AI Consultant in Charleston, SCAn executive buyer’s guide to scoping an AI engagement, evaluating consultants, comparing proposals, and buying a production-ready outcome.Agentic AIWhat Is an Agentic AI System?Agentic AI systems are controller-bounded workflows using models, tools, state, and verification. Learn how they work and when to build one.View all field notes →

Selected public talks & writing

On-demand webinarOntologies and knowledge graphs for tariff strategyOn-demand webinarThe AI Wealth AssistantArticleSemantic architecture and BCBS 239 risk dataArticleFinancial-services data readiness for generative AI

09 · Questions

Questions, answered plainly.

What we do, how engagements are shaped, where work happens, and how to start. If your question is not here, start a conversation and ask it directly.

What does Palmetto Interactive do?

Palmetto Interactive is a national consultancy, based in Charleston, for agentic AI, knowledge graphs, ontologies, data and AI architecture, and practical AI products. Engagements are built for regulated institutions, global enterprises, and other organizations with real knowledge and workflow complexity.

Where is Palmetto Interactive located and what areas do you serve?

Palmetto Interactive has Charleston roots and serves organizations nationwide, with on-site collaboration across major U.S. business markets.

How much does it cost to work with Palmetto Interactive?

We work in four common shapes: Working session, Production Readiness Assessment, Production pilot, Build & operate / advisory. The working session establishes fit before either side commits to delivery. Commercial terms are scoped after the working session and depend on the work required.

What makes Palmetto Interactive different from a typical AI agency?

We connect agentic AI to the knowledge, ontology, architecture, and operating controls that make it useful in production. We match the solution to the business need and carry the question through delivery with measurable outcomes.

Which organizations does Palmetto serve?

Palmetto works best where fragmented data, governance obligations, operational risk, or cross-functional complexity make a connected knowledge layer valuable. Typical engagements include ontology design, knowledge graphs, governed agent platforms, and enterprise delivery programs.

Which industries does Palmetto know?

Financial services, insurance, and manufacturing are where we work most. For banks and asset managers, that's operational resilience, risk, and compliance. For insurers, underwriting, claims, and policy semantics. For manufacturers, field service, parts, and supply-chain intelligence. Selected anonymized examples are available on the work page. They are experience descriptions, not independent verification.

How do I start a project with Palmetto Interactive?

Use the contact form or email [email protected] with the decision, workflow, or knowledge problem you are trying to improve. We review the context and start with a working session before recommending an assessment, pilot, or ongoing engagement.

10 · Contact

Start with the problem.

Tell us where knowledge is trapped, work is repetitive, or a decision is harder than it should be. We will turn that context into a clear next step.

AI that delivers.

Useful knowledge, put to work.

Strategy through production · nationwide

Start a conversation →Review selected work →