Technology Integration for Energy & Utilities in Houston, TX
Technology integration for a Houston utility is almost never a blank-sheet exercise. It's a conversation about why the CIS still can't see a smart meter event in under four hours, why the OMS lost half its trouble call context during the last tropical system, why the GIS and the asset register disagree about which feeders were rebuilt after Uri, and why the integration layer between all of them is a knot of middleware nobody on the current team wrote. The tools inside a Houston-area utility are typically world-class in isolation — Schneider EcoStruxure ADMS, an Itron AMI headend, an Oracle CC&B customer system, Esri's Utility Network on the GIS side — but the seams between them are where the work lives. MSG builds those seams. We treat integration as its own engineering discipline, not a side effect of whatever the last big project shipped. In a post-Uri regulatory environment where PUCT and the utility's own board are asking pointed questions about reliability data, outage restoration times, and DER visibility, the integration layer isn't a nice-to-have — it's the surface the entire reliability story is told through.
Houston context
Houston's utility footprint is unlike anything else on the Gulf Coast. CenterPoint Energy serves the electric T&D across roughly 2.7 million metered premises in the Houston metro, plus natural gas distribution across Texas and a handful of other states. On the generation and retail side, Texas's competitive market means ERCOT is the RTO, REPs handle the billing relationship for most residential customers, and the TDU (CenterPoint) runs the wires and the smart meters. That three-way split — TDU, REP, ERCOT — makes integration in Houston structurally more complex than in a vertically integrated market. Customer move-ins, meter swaps, disconnects, and outage events all have to traverse the TDU's systems and hand off cleanly to dozens of REP systems through ERCOT's 814, 650, and 867 transaction set flows.
Post-Uri (February 2021) and post-Beryl (July 2024), the regulatory scrutiny on reliability reporting has intensified. CenterPoint is spending billions through its resiliency plan — undergrounding, pole hardening, self-healing circuits, additional DMS investments. Every one of those programs pushes more data volume through the integration layer, and every one creates a new integration requirement that wasn't in last year's architecture diagram. ADMS needs real-time DER data that the CIS doesn't naturally carry. OMS needs storm-mode behavior that the current middleware can't scale to during a major event. GIS needs to reflect post-storm rebuild records within a sprint, not the next quarterly sync.
MSG is 79 miles east of downtown Houston on I-10 — about 90 minutes door-to-door to the Energy Corridor, the CenterPoint operations centers off Highway 6, or the downtown regulatory offices near the Allen Center. We're close enough to be onsite weekly during active engagements and close enough that control-system vendors, regulatory counsel, and internal IT leaders can pull us into a working session the same day when something breaks.
Delivery
Every MSG integration engagement starts with a systems audit that goes deeper than the architecture diagram on the wall. We map the actual data flows — not the intended ones — between CIS, OMS, AMI headend, ADMS/DMS, GIS, MDM, asset management, and the billing and collections stack. We read the middleware: what's in the ESB, what's in the iPaaS, what's still running as a nightly SQL job somebody wrote in 2013. We pull sample payloads from real transactions and trace them end-to-end. We interview the people who babysit the integration layer at 2am during a storm.
From the audit we build an integration architecture that's explicit about the IT-OT boundary, NERC CIP scoping, and what data is allowed to cross which zone. We design event-driven patterns where the legacy synchronous flows are causing cascade failures, and we keep the batch patterns where they're working and stable — integration isn't about modernizing for its own sake. We implement against your stack: Kafka or Solace for event transport where it makes sense, MuleSoft or Boomi where the organization already has the muscle, direct API integrations where the vendor finally shipped them, and carefully documented custom adapters where the vendor hasn't. We ship with runbooks, monitoring, and a handoff period where your integration team operates the system with us shadowing, then with us on call, then without us.
Energy & Utilities angle
Utility technology integration has three structural characteristics that most generic integration consultancies underestimate. The IT-OT boundary is real and regulated — NERC CIP-005, CIP-007, and CIP-010 impose concrete requirements on anything that touches the electronic security perimeter around bulk electric system cyber assets, and integration patterns that work cleanly in a bank will get you a finding at audit. Every MSG integration design documents CIP scoping explicitly: what's in, what's out, what compensating controls apply at the boundary, and how change management flows across it.
The AMI-to-OMS-to-CIS loop is where a surprising amount of utility value creation actually happens, and it's where most integration debt accumulates. When a smart meter reports a last-gasp outage signal, the value chain depends on that event hitting the OMS within seconds, enriching with GIS topology data to predict the outage extent, updating the customer-facing outage map, feeding the storm center's crew dispatch, updating the CIS so inbound customer calls get accurate ETRs, and eventually reconciling back to the MDM for billing accuracy. Every broken link in that chain is a measurable SAIDI/SAIFI hit and a customer experience hit. MSG designs this loop as a single system, not as six vendor products.
DER integration is the racing edge. Rooftop solar, battery storage, EV charging load, and aggregated VPP programs are showing up on Houston feeders faster than most utilities' tooling can keep up with. ADMS vendors are shipping DERMS modules, but the integration work between DERMS, the CIS (for the customer-side program enrollment), the AMI (for telemetry), and the GIS (for topology-aware control) is where the actual capability lives. We build that capability.
Why MSG
MSG builds production software for a living. ServiceStorm is a multi-tenant SaaS platform serving home services operators across the Gulf Coast — the kind of system that has to handle real event volumes, real integration surface with third-party tools, and real operational SLAs. MFGBase is a B2B marketplace connecting manufacturers globally, with all the data integration, search, and cross-system workflow that implies. LocalAISource is an AI professionals directory running in production. We know what it means to ship and operate integrated systems because we do it every week on our own products.
That shows up in how we engage with a utility. We don't hand off an architecture deck and leave. We write the adapters. We sit in the control room during the first storm after go-live. We build the observability so your integration team can see what's happening at 3am without paging a vendor. And we scope engagements to end at a system that your people run, not at a dependency on MSG retainer.
We're 79 miles from downtown Houston. That distance matters. Control-system integration work demands real in-person presence during critical phases — we treat Houston as a home market, not a fly-in client. For a CenterPoint-scale engagement, weekly onsite minimum is the default, and we're on site through any major event during the engagement.
Twelve months into a Houston technology integration engagement, you have a CIS that can see an AMI event in under 15 minutes, an OMS that holds context through a major storm without losing trouble call history, GIS topology that matches field reality within a sprint of rebuild completion, and an integration layer your own team can extend without calling a vendor. SAIDI and SAIFI reporting is auditable end-to-end. Customer-facing outage maps and ETRs reflect ADMS state in near real time. DER data is flowing through DERMS-to-CIS-to-AMI with the access controls NERC CIP requires.
FAQ
Our CIS is Oracle CC&B, our OMS is from one vendor, our AMI is Itron, and our GIS is Esri. The middleware is a mix nobody really owns. Where does MSG start?
We start with a two-to-three week audit focused on actual data flows rather than intended architecture. We interview the people who operate the middleware at 2am, pull sample transactions end-to-end from CC&B through the ESB to the OMS and back, and document the real state of the integration layer — including the SQL jobs, the custom adapters, and the undocumented dependencies that every utility accumulates over 10-15 years. From that audit we produce an honest architecture diagram that reflects reality, a risk register of the integration points most likely to fail during a major event, and a prioritized roadmap. Most Houston utilities we talk to are surprised by what the audit surfaces — not because the team is weak, but because nobody has been given the time to look at the whole picture in years. The audit alone is usually worth the engagement before we write a line of code.
How do you handle NERC CIP scoping during integration work?
Explicitly and in writing, from the first week. Every integration we design includes a CIP scoping document that identifies which components are inside the electronic security perimeter, which are outside, and what compensating controls apply at the boundary. We design for CIP-005 (electronic security perimeter), CIP-007 (systems security management), and CIP-010 (change management) from the start — not as a retrofit before audit. Where integration patterns need to cross the ESP boundary, we use data diode or one-way transfer patterns where appropriate, documented API gateways with change-controlled whitelists where bidirectional flow is required, and logging that gives your compliance team the audit trail they need. We've seen too many integration projects do the technical work cleanly and then discover at audit that the CIP story wasn't documented correctly. That's avoidable with upfront discipline, and the documentation produced during the work makes subsequent audit cycles dramatically faster. The auditor isn't starting from cold context.
Post-Beryl, our OMS got overwhelmed during the event and we lost context on thousands of trouble calls. Can integration work fix that?
Partially, depending on what 'overwhelmed' meant. If the OMS application itself couldn't handle the call volume, that's a vendor capacity conversation — we can help scope it but we don't replace the OMS. If the OMS was fine but the integration layer into and out of it (CIS customer context on inbound, crew dispatch on outbound, AMI last-gasp correlation, customer outage map feed) became the bottleneck, that's squarely where integration work moves the needle. In our experience with post-storm reviews, the integration layer is usually where more of the pain lives than the core applications. We'd start by reconstructing what actually happened during Beryl at the integration layer, find the specific choke points, and design storm-mode patterns — asynchronous queues with backpressure handling, degraded-mode fallbacks, event replay capability — that let the next major storm be survivable at the integration level even when one component is struggling.
We have a DERMS project starting and nobody has integrated it with CIS for the program enrollment side. Can MSG own that?
Yes, and it's a common pattern. DERMS vendors ship solid real-time control capability but weak program enrollment and customer lifecycle integration, because the customer relationship lives in the CIS and the DERMS vendor doesn't own that surface. We build the bridge: customer enrolls in a VPP or battery program through the customer portal (CIS-fronted), enrollment flows to DERMS with the right device identifiers and program parameters, telemetry from the device flows back through the AMI headend to DERMS for control and to the CIS for billing, and the settlement data flows to the billing engine. That's a four or five-system integration problem, and it's exactly the kind of work MSG does. We'd scope it as a dedicated workstream inside the broader integration engagement or as a standalone project. DER program economics depend on the full enrollment-to-settlement chain working, and a DERMS without clean integration delivers maybe 40% of its potential value. Building the bridge is usually the single highest-ROI move once a DERMS is live.
We already have MuleSoft/Boomi/an ESB. Do you replace it or work with what we have?
Work with what you have almost always. The integration platform is rarely the actual problem — the problem is usually the patterns being implemented on it, the lack of observability, and the accumulation of undocumented custom flows. We build on your existing platform, using the capabilities it has, and only recommend platform changes if there's a concrete capability gap (for example, high-throughput event streaming on a platform that's fundamentally batch-oriented). When we do recommend a pattern change — say, moving from synchronous request-response to event-driven for AMI-to-OMS — we implement it incrementally, with the old pattern running alongside during transition, and we validate against real storm-mode load before cutting over. Integration projects that start with 'rip out the ESB' are the ones that fail. Platform replacement is a big story that sells well to CIOs and almost never survives production. We'd rather deliver a smaller, measurable integration improvement on your existing platform than sell you a multi-year migration that dies in year two. That's the honest play.
How do you work with our existing control-system vendors during integration engagements?
Collaboratively, and in writing. Most utility integration work touches vendor systems whose data models and APIs we don't get to redesign — Schneider, GE, ABB on the ADMS side, Itron or Landis+Gyr on the AMI side, Oracle or SAP on the customer and asset side. We work with the vendor's published integration interfaces first, engage vendor professional services where the interface doesn't meet the need, and document every deviation. We don't write custom code against undocumented internal APIs because that's a maintenance disaster eighteen months from now. When a vendor needs to be pushed on a capability gap, we help your team build the case — specific use cases, specific data flows, specific business impact — rather than just lobbying. Most vendors engage constructively when the ask is specific and grounded. Vague pressure gets vague responses; specific asks backed by operational evidence tend to move product roadmaps.
Other Industries in Houston
Tech Integration in Other Cities
Other MSG Services
Ready to untangle your Houston utility integration layer?
Let's audit the real data flows, surface the storm-mode risks, and build integration that holds through the next major event.