Ask Quantica · AI case search

Know what an injury is worth,
in seconds.

Ask in plain language: judge, injury, occupation, fatal or not. Quantica reads the question, searches 2,395 citation-safe Hong Kong personal injury and employees’ compensation quantum judgments, 1990–2026, and answers with real cases, every figure pinned to its paragraph.

Request a trialSee pricing

Provisioned trials for law firms, chambers and insurers: preview workspace within one business day.

All cases by Judge Bohwani on accumulation of wealth in fatal personal injury cases
Interpreted asjudge: Bharwaneyfatal accident: yestype: PI
Bharwaney J decided 4 fatal personal injury cases in the corpus. In Kan Wai Ling and Fan Mei Na v Kan Chi Fai [2018] HKCFI 1024 he awarded HK$369,534 for loss of accumulation of wealth, within a total award of HK$4,461,071.
“I award the sum of HK$369,534 as the loss of the net accumulation of wealth in favour of the estate under the provisions of LARCO.”¶ 137 · hklii.hk ↗
4 cases matched
2,395
citation-safe judgments
1990–2026
36 years of awards
372
judges searchable by name, even misspelt
994
occupations indexed
¶-pinned
every figure quotes its source paragraph
Ask Quantica

Plain-language questions. Grounded, cited answers.

Describe the cases you need the way you’d brief a trainee: judge, injury, occupation, court, years, fatal or not. The AI reads the question, searches the corpus, and answers in a sentence or two with the judgments to cite.

  • Misspelt judges understood. Type “Bohwani” and Quantica searches Bharwaney J: every name is resolved against the roster of all 372 judges in the corpus.
  • It knows the concepts. Fatal accident and dependency claims, loss of accumulation of wealth, LOEC refusals, EC assessments: ask in the language of Hong Kong PI practice.
  • Grounded by construction. Answers cite only judgments retrieved from the corpus for your question. If nothing matches, it says so: no invented cases, no invented figures.
  • Transparent, not a black box. Every answer shows how your question was interpreted, as chips: judge, court, years, case type. One click reopens the same query in filter search.

The full bench, one tab away.

Ask Quantica and the classic filter search share one corpus and one results table. Filter by injury, occupation, court, year and outcome, with live medians and CPI adjustment, whenever you’d rather drive than ask.

The Quantica filter bench: injury and occupation search with PSLA medians and LOEC outcomes
The two numbers every claim needs

PSLA by injury. LOEC by occupation.

Hong Kong quantum turns on two figures. Quantica indexes both across the entire corpus, with medians, ranges, and the judgments to cite for each.

PSLA

Pain, Suffering & Loss of Amenities: driven by the injury

1,479 cases with PSLA awards, from HK$100 to HK$2.5M (median ≈ HK$250,000). Search by injury: “rotator cuff”, “crush fracture”, “below-knee amputation”, and read the comparable awards with the judge’s own words.

LOEC

Loss of Earning Capacity: driven by the occupation

468 awards and 422 express refusals. The refusals are a unique asset: reasoned findings you can cite when resisting an LOEC claim, a research answer no headnote service surfaces.

Hours of legal research, answered in seconds.

Quantum in seconds

Filter 2,395 judgments by injury, occupation, court, year, and case type. Medians and ranges compute live over your matching set: reserve-setting and advice letters start from data, not memory.

Citation-safe by construction

Every figure carries the verbatim judgment sentence, its paragraph pin, and the HKLII link:

“I would assess damages for pain, suffering and loss of amenities in the sum of HK$650,000.”¶ 74 · hklii.hk ↗

Refused LOEC as a citable argument

422 cases where the court addressed loss of earning capacity and declined it, with reasons. Defence teams cite them to resist inflated claims; plaintiff teams use them to steer clear of weak heads.

Built for both sides of the claim.

Plaintiff solicitors

Anchor demands in comparable awards. Show the client, and the other side, the median, the range, and the three best cases before the first letter goes out.

Barristers’ chambers

Skeleton-ready citations with the exact paragraph. CPI-adjust historical awards to present-day values, and check whether a case was appealed before relying on it.

Insurers & TPAs

Reserve against the data, not the demand. Employees’ compensation assessments with statutory breakdowns, EC offsets, and nil-after-offset outcomes flagged corpus-wide.

Why not the big databases?

Faster. Cheaper. Defensible.

  • General databases cost a fortune, and charge for the whole library when the quantum question needs one specialist corpus done properly.
  • Headnotes cover only a fraction of cases. Quantica reads every judgment in scope and extracts the award structure itself, including the unreported decisions where most quantum data lives.
  • They are not built for the quantum question. Keyword search returns documents; Quantica answers the question, in plain language if you like, with awards you can put in an advice: PSLA, LOEC, heads of damage, offsets, and net totals.

Frequently asked questions

What is Quantica?

Quantica is an AI-assisted quantum-of-damages research tool for Hong Kong personal injury and employees’ compensation claims. It indexes 2,395 citation-safe judgments from 1990–2026 across the District Court, Court of First Instance, Court of Appeal and Court of Final Appeal, and answers the question every claim starts with: for this injury and this occupation, what is the likely award, and which cases do I cite?

What is Ask Quantica?

Ask Quantica is the AI search inside Quantica. You describe the cases you need in plain language, for example “All cases by Judge Bohwani on accumulation of wealth in fatal personal injury cases”. The AI corrects the judge’s name to Bharwaney against the corpus roster of 372 judges, understands the fatal-accident and accumulation-of-wealth concepts, runs the search, and answers in a sentence or two citing the judgments it found. Every answer shows its interpretation as chips, and one click reopens the same query in the classic filter search.

Are the AI answers grounded, or can they invent cases?

Grounded by construction. The AI may only cite judgments retrieved from the Quantica corpus for your question: it cannot introduce a case or a figure from outside the retrieved set, and when nothing matches it says so rather than guessing. Every figure still carries the verbatim judgment sentence, its paragraph pin and the HKLII link, so you verify before reliance, exactly as with filter search.

What makes the figures citation-safe?

Every monetary figure in Quantica carries the verbatim sentence from the judgment in which it appears, the paragraph number, and a direct link to the full judgment on HKLII. Nothing is paraphrased or estimated: if a figure is shown, you can open the judgment and read the sentence it came from. Extraction accuracy is validated against a hand-checked gold benchmark (PSLA, citation, court and date at 100%).

What are PSLA and LOEC?

PSLA (Pain, Suffering and Loss of Amenities) is the injury-driven head of damages, broadly standardised by injury type, so comparable cases anchor the figure. LOEC (Loss of Earning Capacity) is the occupation-driven lump-sum head under Chan Wai Tong. Quantica indexes both: 1,479 cases with PSLA awards (HK$100 to HK$2.5M, median around HK$250,000), 468 LOEC awards, and 422 cases where LOEC was expressly refused: refusals judges gave reasons for, which are citable arguments in themselves.

Does it cover employees’ compensation as well as common-law PI?

Yes. The corpus holds 1,750 common-law personal injury cases and 613 employees’ compensation assessments under the Employees’ Compensation Ordinance (Cap. 282), with the statutory section 9 / section 10 / section 10A breakdowns, net compensation ordered, and nil-after-offset outcomes flagged.

How is this different from LexisNexis or Westlaw?

General legal databases cost a fortune and their headnotes cover only a fraction of cases, and they are not built for the quantum question. Quantica reads every judgment in the corpus, extracts the award structure itself, and lets you filter by injury, occupation, court, year and outcome, with medians and ranges computed over the matching cases. And you can simply ask: a plain-language question becomes a corpus search with a grounded, cited answer, not a pile of documents to read.

How do trials work?

Trials are provisioned, not self-serve. Request a trial with your firm details and we set up a preview workspace, a 100-case sample of the full corpus with the complete search experience, usually within one business day.

See it on your own cases.

Request a trial and we’ll provision a preview workspace, the full search experience, Ask Quantica included, over a 100-case sample, usually within one business day.

The small-firm plan has a public annual price, so if you’re ready you can skip the trial and subscribe straight from the pricing page.

Request a trialView pricing