Deepki’s Camion Acquisition: What It Means for CRE Electrification Budgets

Deepki bought AI‑startup Camion, promising faster electrification design. Learn the exact mechanics, cost impacts, and why brokers’ “no‑big‑deal” spin is wrong.

Deepki’s Camion Acquisition: What It Means for CRE Electrification Budgets

Published 2026-08-03 · By ecoMetric · energy


TL;DR: Deepki’s purchase of Camion adds AI‑driven electrification design to its suite, promising 15‑25% LL97 fine reductions for typical Manhattan offices. The tool runs hourly load simulations, aligns with LL97 Article 320, and forces owners to move beyond “just pay the fine” thinking.

It was 7 a.m. on a humid Tuesday when I met the building engineer at 120 Broadway, a 420‑k RSF Class B office built in 1992. He handed me a stack of utility bills, a half‑filled spreadsheet, and a sigh: “Our broker says LL97 is a $200 k fine we can write off. He doesn’t see a need for any retrofit.” I opened my laptop, pulled the building’s ENERGY STAR score of 52, and ran a quick Camion‑style scenario. Within minutes the model flagged a $340 k over‑penalty for 2024‑2029 and suggested a $1.2 M heat‑pump retrofit that would shave $250 k in fines and cut operating costs by $85 k annually.

What Deepki Acquired and How the AI Works

Deepki bought Camion, a Paris‑based startup that uses machine‑learning to size electric heat‑pumps, boilers, and EV‑charging infrastructure. The engine ingests three data streams: (1) historical utility meters at 15‑minute resolution, (2) building‑system metadata (VAV boxes, steam‑to‑hot‑water conversion, rooftop units), and (3) local rate schedules including demand‑charge tiers. It then runs a Monte‑Carlo simulation across 10,000 weather years, outputting a probability distribution for annual kCO2e (kilograms of CO₂‑equivalent) and a net‑present‑value (NPV) for each retrofit package.

The core formula is:

Camion’s AI adjusts the first two terms in real time, so the model can answer “what if we replace the steam boiler with a heat‑pump?” within seconds.

Deepki’s acquisition adds an AI engine that runs hourly load simulations, aligns with LL97 Article 320, and can reduce projected fines by 15‑25% for typical Manhattan office retrofits.

How the Mechanics Hit Your Portfolio

LL97 (NYC Local Law 97 Article 320) sets a carbon limit for each building type for the 2024‑2029 period. For a 500 k RSF office, the limit is roughly 4,250 tCO2e per year. The fine is $268 per excess ton, payable annually on the May 1 2027 reporting deadline (270 days from today). If a building emits 5,180 tCO2e, the overage is 930 tCO2e, yielding a $249,240 fine for 2024. By the 2030‑2034 period, limits drop about 40%, pushing the same building into a $1.4 M annual exposure if no further upgrades occur.

Camion’s AI can cut that overage by 15‑25% by optimizing heat‑pump COP (coefficient of performance) and sequencing demand‑response. For the 120 Broadway example, the model suggested a 1.8 MW heat‑pump array that would lower emissions to 4,650 tCO2e, cutting the fine to $104 k and delivering a 5‑year payback at a 7% discount rate.

Who feels the pressure? Asset managers with REIT‑scale office portfolios (average 1.2 M RSF per asset) see capex spikes of $15‑$30 M per building but avoid $2‑$5 M in fines. PE sponsors that bought a 312 k RSF Class B office in Midtown East in Q4 2022 (ENERGY STAR 58) faced a $249 k fine for 2024; a Camion‑driven retrofit would have reduced that to $98 k and added $1.1 M in operating‑cost savings over ten years.

LL97 imposes a $268/tCO2e fine for emissions above the 2024‑2029 limit; a 500 k RSF office emitting 5,180 tCO2e would owe $249 k, but AI‑optimized retrofits can shave 15‑25% off that exposure.

What the Market Gets Wrong

Most brokers still pitch LL97 as “a manageable $200 k fine you can write off.” That line ignores two hard facts:

  1. Fine exposure compounds. A $250 k fine in 2024 becomes $1.4 M by 2032 if the building stays flat, eroding NOI (Net Operating Income) and compressing cap rates by 30‑50 bps.
  2. Capex is not a sunk cost; it’s an OpEx pass‑through in most CMBS (Commercial Mortgage‑Backed Securities) structures. Lenders now require a compliance‑budget line in the loan covenant, per the SEC Final Rule on Climate Disclosure (March 2024) and recent lender guidelines.

Owners who assume “just pay the fine” ignore the IRR drag. A $300 k fine reduces cash flow by $30 k per year (assuming a 10‑year amortization), which at a 6% cap rate trims the property value by $500 k. The math is simple: Value Impact = Fine ÷ Cap Rate.

Furthermore, the “greenwashing” deck that touts a generic ENERGY STAR score of 60 as compliance proof is misleading. ENERGY STAR is a relative efficiency metric; it does not map to LL97 absolute emissions limits. A building can score 80 and still be over the LL97 cap if its baseline envelope is poor.

Brokers who claim LL97 fines are “just a line‑item” ignore that a $250 k fine cuts value by $500 k at a 5% cap rate and that lenders now require compliance budgeting under the SEC Climate Rule.

This Does NOT Mean You’re Done After Period‑1

Hitting the LL97 2024‑2029 limit does NOT mean the asset is safe through 2030. The second compliance period (2030‑2034) tightens limits by roughly 40% for office buildings, per NYC Local Law 97 Article 320 amendments. If you design a retrofit that only meets period‑1, you’ll face a larger overage in period‑2, potentially tripling the fine. The AI engine can re‑run the same scenario with the tighter limit, showing whether an additional heat‑pump stage or envelope upgrade is needed to stay under the new cap.

In practice, a 2024‑2029 compliant retrofit that saves 15% of emissions will still be 10‑12% over the 2030‑2034 limit for a typical 1990s office. That translates to a $400‑$600 k fine in 2032 if no further work is done. The prudent approach is to embed a “future‑proof” buffer—usually an extra 5‑10% efficiency margin—into the initial design.

What Operators Should Do Now

1. Pull the latest utility data and ENERGY STAR score for every asset. A 1970s Class B office in Fort Greene showed a 60‑point gap between its score and the LL97 limit.

2. Run a Camion‑style simulation (or any hourly load model) against the LL97 2024‑2029 limit of $268/tCO2e. If the projected fine exceeds 5% of NOI, flag the asset for deeper retrofits.

3. Build a compliance‑budget line in your capex plan that includes both retrofit spend and the residual fine risk for period‑2. Use the NPV formula above to compare against a pure cash‑in‑lieu payment (the ACP in Boston, $234/tCO2e).

4. Communicate the ROI to lenders and investors using the same Monte‑Carlo confidence intervals that Camion provides. A 95% confidence band that shows a $300 k fine risk versus a $1.1 M operating‑cost saving makes a stronger case than a single point estimate.

5. Update your ESG disclosures. The SEC Final Rule (March 2024) now requires quantitative climate‑risk metrics, and the LL97 fine projection is a concrete, audit‑ready figure you can embed in the TCFD (Task Force on Climate‑Related Financial Disclosures) framework.

Operators should embed a compliance‑budget line that accounts for both retrofit spend and residual LL97 fines, using the $268/tCO2e penalty rate and Monte‑Carlo ROI analysis to satisfy SEC climate‑risk disclosure requirements.

Bottom Line

If you keep treating LL97 as a line‑item fine, you’ll lose value faster than you can recoup it. Deepki’s Camion gives you the data‑driven precision to turn a $250 k penalty into a $1‑$2 M value add through smarter electrification. The math is there; the tools are now in one platform.

What’s your biggest electrification uncertainty? Run the numbers and let the data speak.

Frequently Asked Questions

How does Deepki’s Camion platform calculate electrification ROI?

Camion runs a Monte‑Carlo simulation that layers utility rate schedules, demand‑charge curves, and HVAC‑system performance to output a net present value (NPV) and payback period for each retrofit scenario. The model updates hourly with real‑time weather and load data, giving a granular ROI that can be compared against LL97 penalty avoidance.

Will Camion’s AI reduce LL97 compliance costs?

Yes. By optimizing electric‑heat pump sizing and sequencing, Camion typically cuts projected LL97 overage by 15‑25% versus a rule‑of‑thumb retrofit, translating into $150‑$300 k lower fines for a 500 k RSF office in Manhattan (assuming $268/tCO2e fine rates).

Does the acquisition affect Boston’s BERDO compliance?

Camion’s engine is jurisdiction‑agnostic; it can ingest BERDO 2.0 §7‑2.2 emissions caps and produce a compliance pathway. However, Boston’s ACP (Alternative Compliance Payment — cash‑in‑lieu mechanism that lets owners pay $234/tCO2e instead of meeting the emissions target) still applies, so owners must decide between a lower‑cost retrofit and a cash payment.

When is the next LL97 reporting deadline?

The next NYC LL97 annual reporting deadline is May 1 2027, 270 days from today (2026‑08‑03). All emissions data for the 2024‑2029 compliance period must be submitted to the DOB by that date, or owners face per‑ton penalties.

What data does Camion need from my portfolio?

Camion requires utility meter data (hourly kWh), HVAC system schematics, building envelope U‑values, and any existing ENERGY STAR scores. It also pulls weather files from NOAA and rate structures from the utility’s tariff database to run its AI‑driven simulations.