REDUCTO × ALTER DOMUS
Differentiation Brief · August 2026
Request an evaluation
Prepared for the Alter Domus AI Team

How Reducto differentiates

One extraction layer for the entire document estate, "PDF, Excel or even fax": from everyday invoices and statements to 500-page credit agreements, fund-of-funds structures, bundled multi-invoice files, and K-1s that arrive differently from every preparer. Everyday documents get cheaper and more accurate. The long, bespoke documents become possible at all.

SOURCES ARE MARKED THROUGHOUT: independent benchmarks · competitors' own documentation · Reducto figures. Where a competitor is genuinely strong, this document says so.

Request an evaluation one click drafts the email · or write to mohammed@reducto.ai
long_horizon_accuracy
99.3%

Leaf-level accuracy from Reducto's Deep Extract, with 225 of 225 long documents completed. Every field scored, including fields nested deep inside arrays, at the length where other systems truncate, chunk, or fail. Next-best vendors completed 210 and 203.

micro1 LongExtractionBench · independent · report available on request
recall
99.6%

Recall and precision: the share of fields actually captured. On a credit agreement the expensive failure is the covenant that was never extracted, not the one extracted slightly wrong. Next best: 92.7% and 77.5%.

same benchmark · leaf-level scoring
cost
99%+ · no review

Everyday documents (invoice-class) at 99%+ field accuracy with no human review, read by purpose-built models in the ~100M-parameter class instead of at frontier-token prices. One layer for the whole estate is what moves the cost line.

Reducto claim · verify it on your own documents in an evaluation
01

Your documents, and why they defeat the incumbent stack

Six document families from the estate. The pattern repeats: each incumbent architecture was built for a different document than the one in front of it. The grids in 02 and 03 score these same failure modes vendor by vendor.

Invoices & everyday forms · 99%+ without review Capital-account statements & notices Excel-native loan tapes Faxed & scanned notices Schedules of investments K-1s, every preparer Fund-of-funds documents Bundled multi-invoice files Credit & loan agreements · 50 to 500+ pages

Credit & loan agreements

50 to 500+ pages · amendments layered on top

One agreement, several amendments. The current value of a covenant or pricing grid lives in whichever amendment last touched that section, and defined terms cross-reference each other across hundreds of pages. The document is a legal argument, not a form.

Template IDPfree prose has no layout to anchor a template to
OCR servicesreturns words and boxes; "which amendment wins" is not an OCR question
Platform extractDatabricks errors past 500 pages, from its own docs; BDA has no public long-document evidence
Frontier directattention degrades past ~100 pages; truncation is silent
ReductoDeep Extract completed 225/225 long documents at 99.3% leaf accuracy; amendment supersession is instructable in the schema

Fund-of-funds documents

sections inside sections · the schema is a tree

The same fields repeat per underlying fund, per share class, per period: arrays inside arrays. Flat key-value extraction destroys exactly the hierarchy the accounting needs, so the output is unusable even when every individual value is right.

Template IDPzonal templates are flat by construction
OCR serviceskey-value pairs with no hierarchy
Platform extractBDA blueprint fields are string, number, or boolean, per AWS's docs
Frontier directnested JSON degrades as depth and length grow
Reductodefine the array once; structure survives to the last sub-section

Bundled invoice & statement files

dozens of documents · one PDF · no separators

Intake arrives as one blob. Each invoice or statement spans two to three pages with no separator sheets, no barcodes, and no fixed boundaries, and each piece needs its own extraction before anything downstream can run.

Template IDPseparation wants barcodes, separator pages, or hand rules
OCR servicestrained custom classifiers where offered; nothing out of the box
Platform extractparses the blob; boundaries stay your problem
Frontier directmanual orchestration, brittle at volume
Reductosplit by plain-language description; each piece routed to its own schema

K-1s

a different layout from every preparer

Thousands of preparers, each rendering the federal K-1 its own way, with the values that matter pushed onto attached statements: box 20Z becomes an arbitrary-length array living on attachment pages that no prebuilt model has seen.

Template IDPa template per preparer is economically impossible
OCR servicesprebuilt tax models cover common forms, not the attachments where the values live
Platform extractblueprint tuning per variant, examples in hand
Frontier directworks until the attachment format changes
Reductoone schema, zero-shot across preparers; attachments captured as array fields

Capital-account statements & notices

hundreds of GP layouts · "PDF, Excel or even fax"

Every GP formats commitments, contributions, distributions, and NAV its own way, and the estate arrives in mixed media: born-digital PDFs, Excel-native files, and scanned or faxed notices with real-world artifacts.

Template IDPper-GP layout drift breaks templates silently
OCR serviceshandles the scan, not the bespoke semantics
Platform extractDatabricks' docs warn on dense or low-resolution content
Frontier directvision quality varies on degraded scans
Reductoscan-hardened parsing, Excel-native input, zero-shot schemas across layouts

Schedules of investments & waterfalls

tables that fight back

Merged cells, no gridlines, footnoted values, and tables that continue across pages. One misplaced cell silently shifts a column of valuations, and the error survives every downstream check because the table still looks like a table.

Template IDPper-template table zones, rebuilt on every change
OCR servicesone generic pass; merged cells shear
Platform extractHTML out; accuracy unbenchmarked on tables like these
Frontier directloses bidirectional row and column referencing
Reductomulti-pass checks over rows, columns, merged and empty cells; best model routed per table
02

Feature by feature: capability

  yes   partial, or with material caveats   no
Capability ABBYY Hyperscience Textract / Azure DI Bedrock Data Automation Databricks ai_parse_document Frontier LLM direct Reducto
New layout, zero setup: a K-1 from a new preparer, a lender format never seen before template first labeled samples first generic fields; custom needs training1 blueprint + 3–10 tuning examples2 parse only; extraction is a separate generic LLM call3 prompt-dependent zero-shot, schema or schemaless
500+ page agreement, one shot, no chunking OCR yes; document-level extraction no no public long-document evidence hard cap: 500 pages / 100 MB3 truncates and degrades silently no chunking penalty; Deep Extract completed 225/225 long documents4
Amendment-aware values: the third amendment supersedes section 4.2 per-field instructions exist; long-document reasoning is the gap can reason, unreliable at length instructable in the schema
Nested schemas: fund-of-funds arrays inside arrays blueprint fields are string / number / boolean2 fragile as depth grows define the array once; it iterates
Bundled files: dozens of 2–3 page invoices or statements in one PDF rule / barcode separation trained classifier trained custom classifier (Azure); nothing zero-shot manual orchestration split by plain-language description, routed to per-type schemas
Merged-cell, gridline-free, footnoted tables: schedules of investments, waterfalls per-template zones one generic pass HTML out; accuracy unbenchmarked3 loses row / column integrity multi-pass structure checks; best model routed per table5
One API across the estate: born-digital PDF + Excel-native + faxed scans strong on scans; Excel via config docs warn on dense or low-resolution content3 vision quality varies all three, one call

1 Vendor docs: Textract Queries / Azure custom models require configuration or training per document type.   2 AWS BDA docs and blog; AWS also calls schemas "blueprints," not to be confused with internal blueprints.   3 Databricks' own function docs: 500-page / 100 MB limits, English-tuned, "may take a while to run or ignore content."   4 micro1 LongExtractionBench, independent, 225 long documents; report available on request.   5 When a frontier model beats Reducto's own on a table type, routing surfaces it automatically; continuous evaluation decides.

03

Feature by feature: the operating model

Operating question ABBYY Hyperscience Textract / Azure DI Bedrock Data Automation Databricks Frontier LLM direct Reducto
Per-field confidence usable for auto-accept OCR / rule-level automation-rate thresholds, built on trained samples field confidence on prebuilt / custom models; generic output is word-level broad bands: ≥90 high, 70–89 medium6 none for extraction token probabilities are not calibrated per extracted field; see 04
Citation tied to each extracted schema field zones, not fields on prebuilt / custom models; generic output is words + geometry per field element metadata, not extracted fields per-field bounding box + page
Setup for a new document type build a template label and train configure or train a custom model author a blueprint + examples write prompts around the parse write prompts + build evals paste an existing JSON schema
When layouts drift or models change, who fixes accuracy you you (retrain) you shared you you (the prompt treadmill) Reducto: continuous evaluation, auto-optimization, shared engineering channel
Independent third-party evidence on long-document extraction not evaluated7 not evaluated7 not evaluated7 not evaluated7 not evaluated7 n/a micro1: 99.6% recall / precision, 99.3% leaf accuracy4
Deployment on-prem ✓ on-prem ✓ their cloud their cloud (AWS) your lakehouse API / your Bedrock SaaS, hybrid VPC inside your AWS, or on-prem
Throughput at a 30M-docs / month scale capacity licenses capacity licenses scales scales batch rate limits + retries 1,000 req/s at the edge; batches queue rather than reject

6 AWS's published guidance brackets confidence into bands and recommends human verification below them (AWS blog).   7 "Not evaluated" means not present on the benchmark cited; each vendor can be asked for its own independent long-document results.

04

Confidence per field, citations per field

Every extracted field carries two properties: a confidence score you can threshold, and a bounding-box citation to the exact page region it came from. Together they set the only number that matters operationally: how much of the estate flows through with no human review at all.

Illustrative output · credit agreement, 512 pages · synthetic document
borrower_name "Aldgate Continental LP" 0.99 p.1 auto-accept
facility_commitment "$250,000,000" 0.98 p.14 auto-accept
maturity_date "2031-06-30" 0.97 p.9 auto-accept
applicable_margin_bps "275" 0.61 p.412 · Amendment No. 3 route to review
Illustrative values. The routing rule is yours: fields at or above your threshold flow straight through; the rest arrive in a review queue with the citation already open, so a reviewer verifies against the highlighted region instead of re-reading the document.

The threshold is the business case

Set an acceptance threshold, and review effort stops scaling with volume and starts scaling with genuine ambiguity. Everything above the line needs no reviewer; everything below arrives as an exception with evidence attached.

Calibration is the test

A confidence score is only useful if 95 behaves like 95. In an evaluation, plot confidence against actual correctness and find the threshold where auto-accepted fields meet your error tolerance. That one chart converts extraction accuracy into staffing math, for any vendor.

Mapped to evaluation criteria

Groundedness: every field cites the page region it came from. Faithfulness: superseded and struck values are never invented, and the citation proves it. Accuracy: scored per field, nested fields included. For the highest-stakes fields, the Deep Extract harness adds verification passes, the mode behind the 99.3% / 99.6% numbers above.

05

Where the cost difference actually comes from

Document AI cost is four lines, not one. Reducto moves all four.

1 Model spend
Small models read; frontier models reason

Purpose-built models in the ~100M-parameter class handle layout, OCR, and table structure, with frontier models reserved for the cases where they genuinely win. You stop paying frontier-token prices for document reading, and the models you do run work over clean structured text instead of raw PDFs.

2 Human review
99%+ invoice-class accuracy, no review

Per-field confidence plus per-field citations let high-confidence fields flow straight through and route only genuine exceptions to people.

3 Engineering maintenance
Zero templates, training sets, or prompt re-tuning

No templates to rebuild, no samples to relabel, no custom models to retrain, no prompt-and-eval treadmill when a new model ships. Chasing the frontier is Reducto's product, not a standing item on the team roadmap.

4 The cost of what was missed
99.6% recall at leaf level

A covenant, fee term, or amendment that was never extracted becomes rework, breaks, or client exposure downstream. Recall is the quiet budget line that never shows up on a per-page price sheet.

Per-page price comparisons capture line 1 only. Lines 2 through 4 are where document programs actually spend, and they are the lines this architecture was built to attack.
06

Vendor by vendor, in one breath each

ABBYY

Strong heritage OCR on stable scanned forms; every new GP or lender variant is a template build, and layout drift breaks silently.

Hyperscience

The best of trained IDP, and its automation-rate framing is right; it reaches those rates through labeled training samples per document class, which becomes the product you operate on an estate with hundreds of drifting formats.

Textract / Azure Document Intelligence

Excellent, cheap OCR infrastructure returning words and boxes; finished schema fields, splitting, validation, and evaluation remain your build.

Bedrock Data Automation

The serious hyperscaler option, with blueprints, confidence, and citations; a single generic model stack, confidence guidance that recommends human verification below 90, and no public long-document evidence.

Databricks ai_parse_document

Convenient bulk ingestion where the data lives; a parser, not an extraction system, with a documented 500-page / 100 MB cap that the longest agreements exceed outright.

Frontier models direct

The right reasoning engines and the wrong document readers past ~100 pages; give them structured input and they do their best work, which is why Reducto complements rather than replaces them.

07

How it fits, and terms

Modularity

Parse, Split, Extract, and Edit are independent APIs, callable from Databricks pipelines, Bedrock agents, LangGraph, or Claude tool use and Skills.

Existing JSON schemas work as-is; existing orchestration and business logic stay untouched.

Security

SOC 2 · GDPR · DPA included in the trial agreement · zero-data-retention configured on the account.

Documents are never used for training, so evaluation can run on real client documents rather than redacted mockups.

Deployment

SaaS, hybrid VPC (processing inside your AWS while Reducto manages GPUs and model updates), or on-prem.

EU-stringency requirements mapped against our security attestation on request.

Test every claim here on your own documents

Ten documents per schema is enough to separate the field. Reducto provisions the environment and credits under the trial terms; your team keeps the scorecard, judged on your own criteria: faithfulness, groundedness, accuracy.

The click opens a pre-addressed email; add your document types and send.