The AI platform for weather, risk, and operational intelligence

Enter a location. Let the Agent build the console. Launch a command center overnight.

OpsCon turns a simple conversation into a working command center. It brings live information onto a map, sends alerts to the right people, and helps create reports based on what is happening. Build it for your clients or for the organization you work for. No coding needed, and you can change it as your needs change.

Builder· Batangas Terminal GroupDemo 1 of 3
Set up a command center for Batangas Terminal Group: three berths and a depot, rainfall, lightning and typhoon tracks, thresholds per site, a duty roster of six, wearing their own brand.
  1. Four sites pinned on the Batangas coastlocation
  2. Rainfall, lightning and cyclone layers switched onlayers
  3. Threshold rules written per hazard, per siterules
  4. Duty roster of six, escalation ladder setroster
  5. Batangas Terminal’s mark, colours and sign-in address appliedbranding
Live. The duty officers sign in at their own address.5 days · demo
Use OpsCon in two ways

Build a command center, or run intelligent assessments

Choose what fits your work. You can build it for your clients or use it for the organization you work for.

Build a console for monitoring, forecasting, early warning & response

The full route, step by step Playing
Step 1 of 4

Bring the data in: upload a file, connect a live feed over an API, let the AI find a public source, or use a connector. Your thresholds and SOPs are codified here too.

Read this step →

Quicker route: Builder → Model → Deploy, when the hazards are the usual ones and nothing new has to be connected first.

1 Step 1 of 4 · Knowledge

Bring in the information you need.

Upload your files, connect live data, or let AI Agents find public information for you. You can also connect ready-made data sources.

Add your own rules, procedures, important numbers, and steps to follow. Check them once, then use them again and again.

Upload a fileAPI & connectorsLet the AI find a source
Knowledge · sources Knowledge Agent
Uberth-thresholds.xlsx · uploadedFile
APAGASA rainfall · connected over APILive feed
Global lightning · found for you by the agentPublic
CTerminal SCADA · connectorClient
Codified24 of 24
Datasets4 sources Layers9 available Parameters5 watched Rules12 thresholds
The agent’s mapping · a person approves every row
Rainfall mm/hLayer + threshold
Berth closure SOPEscalation ladder
2 Step 2 of 4 · Builder

Build your own console.

The Builder Agent helps you assemble your console. Just talk to it about the parameters you want to monitor and forecast, the tables and graphs you want to read, and the GIS layers you need on the map. It builds the console around what is important to you.

Your sites go on the map, the feeds come in, the alerts and the people responsible are set, and the reports are ready. You can change anything later. You do not have to set everything up one piece at a time.

BuilderAny location on landModules & general features
YouBatangas terminal, four sites. Watch lightning, weather, tropical cyclones and air quality.
4 parameters matched · 9 GIS layers in range
Lightning Weather Tropical cyclone Air quality
Warning thresholds
Lightningred under 15 km
Rainfallred over 30 mm/h
Cyclone windred over 118 km/h
GIS activated · live operating picture
+24h +48h TY BASYANG Berth 1 Depot Berth 2 Berth 3
Rain radar Cyclone track Lightning
Now+12h+24h+36h+48h
2 Step 2 of 4 · Builder

Set up the console for your site.

The Builder Agent helps you assemble your console. Just talk to it about the parameters you want to monitor and forecast, the tables and graphs you want to read, and the GIS layers you need on the map. It builds the console around what is important for that site.

For example, for a solar plant, you can monitor how much power it should produce and what could affect it, such as storms, flooding, heat, lightning, or poor weather.

Everything is set up for the location and site you choose. You can change, remove, or add anything you need.

BuilderGeneration & Hazard modulesEvery module editable after
YouA 113 MWdc solar plant at Location 08. I need the generation forecasts and the hazards that actually reach a plant here.
Location pinned · 4 modules proposed
Solar power forecast458 MWh expected tomorrow
Long-term performance682 GWh over 2027–2031
SoilingClean in 5 days, or lose 3.1%
Typhoon hazardTY2026-14 · arrival in 31 hours
Publish app
Solar Plant · Location 08113.8 MWdc · LIVE
GIS · plant and hazard
LOCATION 08
Forecast · MWh
Today412P50-8%
+1d458P50ok
+2d331P50cloud
+3d96P50typhoon
3 Step 3 of 4 · Model

Calibrate it against the real operation.

Once the console is built, you manage and calibrate it. Choose the data each site shows, such as rain, weather forecasts, lightning, storms, wind, satellite information, or your own site data, and switch the GIS layers on.

Then set the rules and the limits that matter, name the recipients, and tune the configurations that count. When something reaches a limit, the right person gets an alert by text, email, and in the console.

If they do not respond, the alert can automatically go to the next person. Everything is recorded, so the console stays aligned with the real world operation.

GIS layers & live feedsThresholds & escalationAssets plotted against what is coming
GIS · Site 23
15 KM Site 23
Base map Lightning · every 15 min
Lightning datafetch frequencySaved
1 h30 min15 min5 min
New alertArmed
WhenLightning inside the 15 km ring
Send toofficer@site23.com
ChannelSMS, email and the console
3 Step 3 of 4 · Model

Calibrate it, then model anything.

Once the console is built, you manage and calibrate it: the data and the rules behind every reading, the recipients, and the configurations that matter, so what you see stays aligned with the real world operation.

Then ask what you want to know in simple words. AI Agents can analyze the information for a specific site, whether the job is simple or complex, combining weather, site information, past data, and other sources to produce a more useful result for that location.

For example, it can estimate tomorrow’s solar power, check flood or storm risks, or look at how conditions could affect a site.

You can use different models and data sources, and check how well they perform before using the results.

ECMWF ensemble · WRF downscalingAI correction on site historyBacktested · P10 / P50 / P90
Model run · Solar PlantRUNNING
Capacity113.8 MWdc80.1 MWac
Array248,000modules, single-axis
Horizon48 hhourly, day-ahead
ECMWF ENS9 km · 51 members WRFnested to 1 km AI correction26 mo. site history Plant ACP10 / P50 / P90
Run log
ECMWF ensemble retrieved51 members
WRF nested to 1 km over the site48 h
GHI → POA, tracker geometry & shadinghourly
Cell temperature, DC losses, soiling−6.1%
Inverter clipping at 80.1 MWacrunning
P10 / P50 / P90 assembledqueued
Method chosen by backtest · 30 held-out days, lower is better
NWP only7.9%
AI only5.2%
Blend4.8%
Modelled output, tomorrowP50 with the P10–P90 band
P50 tomorrow458 MWh
P10–P90426–490
Peak63.8 MW
4 Step 4 of 4 · Deploy Command Center

Deploy your command center.

What you deliver is a real, working command center, not just a report or presentation. It can have your own name, logo, colours, and web address. Your team gets the map, alerts, reports, and tools they need in one place.

You decide who can access it, what they can do, and how much AI Agent credit they can use.

AI Agents work inside the console, monitoring information, checking your rules, identifying what could be affected, and helping your team respond. You manage everything in one place.

Live command center · their brandRoles, roster & AI creditsManaged from the Partner Portal
Batangas Terminal Operations LIVE02:14 PHT
+12h +24h TY BASYANG Berth 1 Depot Berth 2 Berth 3
Rain radar Cyclone track Lightning Flood extent
0 – 50 km
Berth 2 · rainfall 32 mm/h, over your 30 mm/h rule R. DELA CRUZ · ACK 94s
Rainfall forecast · mm/h, per site
SiteNow+3h+6h+12h
Berth 1614198
Berth 23238219
Berth 3491715
Depot271613
NormalWatchWarning · your rule fires
4 SITES128/131 ASSETS ONLINE1 ACTIVE ALERTRADAR 4 MIN AGO
4 Step 4 of 4 · Generate Analysis & Reports

Create the report you need.

Turn the results into a report your clients or team can actually use. Use your own format and branding. The report can be a forecast, assessment, situation report, or any other analysis you need.

A person reviews and approves the report before it is sent out. Then the same console is ready to run again tomorrow, or for your next site.

Day-ahead forecasts, SitReps & assessmentsYour brand on the coverReviewed & released before it goes out
Solar ForecastDay-ahead generation for nomination and trading
LOW RISK
458MWh

Expected generation tomorrow, Sunday 13 September, at ±7% (P10–P90). Market nomination closes 20:00 today.

Forecast band426–490MWh · P10–P90
Peak output63.8 MW12:00–13:00
Capacity factor23.8%vs 22.4% Sep norm
Model error4.8%nMAE, 30 days
Hourly shape113.8 MWdc / 80.1 MWac peak 63.8 MW 6:008:0010:0012:0014:0016:00
Why the number moved
Thunderstorm cloud cover, 11:00–15:00−37
Below-normal GHI in the ECMWF ensemble−19
Cooler ambient after rain, +0.9% efficiency+7
Soiling carried from today−10
Next daysConditionsMWhBand
Sun 13Thunderstorm458±7%
Mon 14Overcast421±9%
Tue 15Partly cloudy509±11%
Wed 16Partly cloudy514±13%
Thu 17Light rain457±16%
Reviewed by M. Ravelo · released 16:20Export PDF
Become a partner

Your first command center is one location away.

Sign up free, build a demo in an afternoon, show it to the operations room, and upgrade it the day they say yes.