Head-to-head · updated 13 September 2026
DataShield vs Syniti: do you need master data management, or do you need an SAP migration?
Syniti has moved SAP data for decades. The Knowledge Platform bundles six modules. Migrate, Quality, Match, Replicate, Catalog & Governance, Master Data Management. It's an SAP Endorsed App on the SAP Store. Their page claims knowledge from more than 5,000 global migrations. Capgemini's release says the deal added more than 1,200 data specialists. That's real muscle for a cutover. We won't pretend otherwise.
Ours is not a migration tool. We build master data for agents. An agent can interview your business users. It derives a spec. It walks a choice gate with you. One step then builds the whole draft model. The matching underneath is Fellegi-Sunter linkage. You can check it, train it, or roll it back. No config reaches production without a green paired-gate check. A golden record must clear a quality gate first. Only then does an agent see it. Consent is a fail-closed veto. We tokenize data at ingest. Agents query tokenized data over MCP. Detokenizing is a special, audited step. Below is the honest split. It names the rows they win outright.
The short version
Pick DataShield when
- You want the model built with you, not sold to you. An agent interviews a business user. It derives a spec. It works an ordered choice gate. Each proposal carries its what, why, impact, alternatives. Override one. Each linked choice flips back for a fresh look. One step then builds the model. All of it draft. How Ontology does it.
- You want to see the match method. Ours is Fellegi-Sunter linkage: fuzzy, not a black box. It uses named comparators: Jaro-Winkler, Levenshtein, Soundex, Double Metaphone. Plus LSH blocking. Plus EM-trained weights on a frozen holdout. Nothing gets promoted without a paired-gate check on the real scoring path.
- Agents are going to read the golden record. Ours have to earn that first. A composite quality score covers six axes. Each axis must clear its floor. Consent is a fail-closed veto. Reads come through views that mask in the view, not in a setting. Each call lands on an audit chain. You can check it yourself. Run the verifier.
- You want your own team to run this next quarter. Self-hosted. At a price you can read on a page.
Pick Syniti when
- You are moving ECC to S/4HANA. This is the thing they do best. It's a genuinely hard job. We don't do it at all.
- You want the software and the people from one place. Capgemini's own release says the deal added more than 1,200 data specialists. They can staff a multi-year programme in twelve countries. We cannot.
- SAP wants an Endorsed App on the SAP Store. Your board wants that box ticked.
- You need Migrate, Quality, Match, Replicate, Catalog and MDM under one contract. Across many domains. With screens your plant stewards will actually use.
Bottom line: They sell a migration programme with MDM attached, plus people to run it. We sell a mastering platform an agent can stand up with you. Proof comes with it. If S/4HANA is the project, call them. Is that project done? Are you just renting a platform for three domains now? Call us.
Feature by feature
Competitor cells describe Syniti's public site and Capgemini's press releases. As of the date above. If we got it wrong, email support@myorg.ai. We will fix it, credited.
| What matters | DataShield | Syniti | Edge |
|---|---|---|---|
| Standing up the model | An agent does it with you. The mdm_project surface is 15 commands. It interviews a business user. It derives a spec. It presents an ordered choice gate. Each proposal carries a what, why, impact, alternatives. You confirm or override. An override propagates to each linked choice. There is deliberately no accept-all. One step then builds the entity types and hierarchies. It also drafts match configs, reference bindings and rules. Each object lands with an honest gate state. The tool never promotes a config. It never runs matching. It never skips the check. | A method, a platform, people. Standing up a domain is a project. Templates, workshops. A partner team sets up the modules. That's the model they sell. It's the one their clients bought. | ◆ |
| SAP S/4HANA migration | Nothing. We have no migration tooling. No SAP add-ons. No opinion on selective data transition. | The core business. Advanced Data Migration and Management by Syniti. An SAP Endorsed Application, an SAP Store listing. A method their platform page describes as drawn from more than 5,000 migrations. | ◇ |
| Match and entity resolution | Fellegi-Sunter fuzzy record linkage. Named comparators: Jaro-Winkler, normalized Levenshtein, Soundex, Double Metaphone, n-gram, date and numeric proximity. LSH blocking with a bucket-size cap. EM-trained m and u weights on a frozen holdout. A flipped signal is a hard refusal, not a warning. PSI drift checks watch each live config. It only proposes. It never acts. An absolute-ID tier routes ID clashes to review. Each gets tagged with its clash class, not merged. A trial writes nothing. Promotion needs a fresh green paired-gate check. | The Match module. Proprietary algorithms and natural language processing, named on the platform page. The method itself is not public. You can't check the score. | ◆ |
| Golden records and survivorship | Five set survivorship rules: most trusted, most recent, longest, most frequent, aggregate. Each has a default chain and attribute groups. Plus a per-value origin record. Most trusted runs on measured source trust. That's tracked from how often a source's calls get overturned. It is not a hand-ranked list. The version log is append-only. It survives deletion. You can rebuild a record as of a date. Merge, unmerge, separate, defer, link: all first-class. Plus record locks and field pins. Rules gate each release. | Master Data Management module. Covers customer, product and vendor domains. Workflow and stewardship screens included. Survivorship rules exist. The strategy set is not public. | ◆ |
| Multidomain modelling | An open entity-type vocabulary. You model your domain rather than bend a template. Three wired matching modes per type. A People, Places and Things taxonomy. Held as data, not a fixed list. Parent-scoped hierarchies. A child config resolves inside its parent's scope. That uses a base threshold. Cross-parent, it resolves at a stricter one. Relationships carry two-sided time validity. Domains carry an owner and a compliance framework. What we don't ship: a boxed Customer 360 or Product 360. We won't pretend our product-module ID checkers are a PIM. | Customer, product and vendor domains out of the box. Plus years of module breadth around them. Plus screens plant stewards already know. Want a boxed domain model on day one? This is the side of the table it sits on. | ◇ |
| Ingestion and binding | Six on-ramps. Inline, URL and batch ingest. Storage watch sources. A file repository. A catalog bind. The last binds a live PostgreSQL source. No rows leave it. Bind a table to an entity type once. Mastering then stays current on a set cadence. It runs under one rule: auto, agent-review or steward-review remaster. A weekly full pass reconciles deletes. Differential loads mark deleted source records stale. They flag updated ones for a fresh look. A delta run refuses. It won't quietly balloon into a full scan. We are not an ETL product. We won't sell you one. That's PostgreSQL today, not your whole estate. | Replicate, Migrate and smart metadata scanning. Connector reach across SAP and the systems around it. Much broader source coverage than ours. The remaster and review question isn't answered in their public docs. | ◈ |
| Stewardship and earned automation | Stewards review patterns, not items one at a time. Each proposal carries evidence, plus a counter-check. A three-rung trust ladder counts human calls, not disposed items. 50 approvals at a 98% rate with zero dissents promotes a pattern. One dissent demotes it. At the top rung, a fixed sampler pulls 10% of auto-disposed items. Those go to human audit. You can prove why any item was picked. A standing rule is born staged. It needs a second clear release. It auto-suspends on an audit overturn. Blast-radius dry runs call the writer's own gate code. A preview cannot say yes and then refuse. | Workflow, approval and stewardship screens built for business users. In front of real plant and finance staff for years. Ours are plainer. What we could not find: a public trust model. Also missing: an audit sample rate. And a rule that stops itself when a steward overturns it. | ◆ |
| MCP and agents | 143 MCP commands for mastering alone: mdm 99, stewardship 29, mdm_project 15. That's out of 447 on Ontology. Roughly a third of the agent surface is master data. An agent can propose a match. It can work a stewardship queue, or enact a rule. It can also read a masked golden record. That's under a scoped tool token. | We found no MCP server on syniti.com. No agent tooling. No tool-calling language, either. Their 2026 copy talks about an AI-ready base. Embedded AI, smart metadata scanning. That's AI in the pipeline, not agents at the door. | ◆ |
| Agent authorization and admission | Two gates, not one. Each checked tool call passes a scope ceiling. It checks an access tier and a revoke check. That happens before dispatch. It fails closed. Delegation is RFC 8693 token exchange. It carries an enforced scope ceiling. Then the record itself has to qualify. A golden is agent-ready only if it clears the bar. That's a composite score. Six quality axes in all. Each axis sits above its floor. Consent must validate too. One rule enforces that. It covers the SQL path. And the app path. | Platform roles and workflow approvals for human stewards. Nothing agent-specific in their public docs. No idea of a record earning the right to be read. | ◆ |
| Audit evidence | A SHA-256 hash chain with Ed25519-signed checkpoints, themselves chained. Checking it tells tampering, insertion, deletion and truncation apart. Each in-place oversight change writes a before and after image. Same step, no gap. Merges, unmerges, pins and overrides are all on the record. Try the verifier. | Change log and workflow approval records, inside the platform. We found no public tamper-proof method. | ◆ |
| Break-glass | Scoped, time-boxed emergency access for agents and operators. It auto-revokes and cannot be quietly deleted from the log. | Not described in their public docs. | ◆ |
| GDPR erasure | Crypto-shred of per-subject key material, plus ISO 27560 consent receipts. A consent record tracks a lawful basis. One per subject and purpose. That covers all six GDPR Article 6(1) bases and nine purposes. No recorded basis means no. Only the derived consent score crosses the masking line. The audit chain still verifies afterwards. | Data oversight and quality tooling that can find and fix records. The erasure method isn't described. Neither is what happens to the log. | ◆ |
| Tokenization and masking | Deterministic, join-preserving, vault-reversible tokens at ingest. Plus quasi-identifier generalization: dates become year, decade or age band. ZIPs become 3 or 4 digits. Phones, SSNs and emails get partial masking too. Each column gets a measured cardinality-reduction score. These are features you switch on, not defaults. Golden consumption views mask in the view. The reader role has no grant on the base table. The mask holds even from SQL. Matching runs on real source values. A tokenized matching field makes the run refuse. It won't mega-merge. | Data quality rules and masking in migration contexts. Tokenization is not a word they use on the platform page. | ◆ |
| Classification and catalog | 129 field classes cover PII, PHI, financial data and secrets. Deterministic and reproducible from a config digest. Covers all 18 HIPAA Safe Harbor IDs. In-place scan and profiling of a live PostgreSQL source. No rows leave it. That's PostgreSQL today, not your whole estate. | Catalog & Governance module, plus smart metadata scanning. Connector reach across SAP and the systems around it. Far broader source coverage than ours. | ◇ |
| Deployment | Self-hosted, in your own cloud or data center. Or a dedicated single-tenant server we run. Docker images ship for Auth, Ontology, Corpus and Lighthouse. Guardian checks each image against a signed deploy manifest. Ed25519 audit-signing keys stay on your instance. You control the key directory. Rotation is kid-keyed. It fails closed on an unknown key. KMS or HSM custody is a built seam in Auth. It's not a shipped link. HMAC tokenization keys sit in your setup today. Not in a KMS. | Described as an all-in-one cloud-based platform. With on-prem roots in SAP-adjacent estates. Current cert list wasn't confirmable from public pages. | ◈ |
| Services and ecosystem | Small team, design-partner terms, source escrow. You run it. Guided onboarding does a lot of the first pass. There is no army. | Capgemini. Their release claims more than 1,200 data specialists from the deal alone. Plus global delivery. Plus named programmes with Bridgestone, Phillips 66, Zurich and Exxon Mobil. | ◇ |
| Maturity signals | Lineage back to 1996: we don't have that. Auth, Ontology, Guardian and Lighthouse run live now. Guardian and Lighthouse, since April 2026. The mastering engine gets checked against roughly 370,000 real public records. Across three City of Chicago datasets. We publish no record-count ceiling. SOC 2 is planned, not attested, and we say so. | Founded 1996. Rebranded Syniti in 2020. Bought by Capgemini in December 2024. Still shipping quarterly (a Q2 2026 release is promoted on their homepage). | ◇ |
| Pricing | Public model, scoped instant quote, no sales wall. | Quote only, through the Capgemini channel. One third-party estimate: about $150K to $350K a year. That's for one or two modules, mid-market. They publish no figures. Treat that as an estimate and ask them. | ◆ |
◆ DataShield leads◇ Syniti leads◈ comparable
Their claims are drawn from syniti.com. Plus the SAP Store listing and Capgemini's press releases. Last checked 13 September 2026. We link them below rather than work from memory.
Three things you get here that you won't get from an enterprise MDM suite
An agent builds the model with you
A suite sells you a domain model. Plus a team to fit it. Here an agent interviews the person who knows the business. It derives a spec. It walks you through a choice gate. One call at a time. Override one answer. Each choice that depended on it comes back for review. One step then builds the whole model. All of it lands as drafts. Drafts still have to earn promotion. See Ontology.
Authority that can change mid-flight
A contractor finishes on a Friday. Their agent is 20 minutes into a 40-minute stewardship run. The next checked tool call gets re-checked. It's checked against current access. It fails closed. No waiting for a token to expire. How Auth does it.
Proof that survives an audit
A change log inside a platform proves what the platform says happened. Ours is a hash chain, with signed checkpoints. The verifier names what broke: tampering, insertion, deletion or truncation. That's the property EU AI Act Article 12 and HIPAA §164.312(b) reviewers care about. Try it in your browser, no signup.
Where Syniti is genuinely stronger
Start with the obvious. Say your data problem is an S/4HANA cutover. Syniti has done it hundreds of times. Few firms on earth have. The method is baked into the tooling. Their page claims knowledge drawn from more than 5,000 global migrations. You can't code around that in a sprint. The Knowledge Platform is an SAP Endorsed App. That matters to architecture boards. Engineers find it annoying. Buyers find it decisive. Add Capgemini behind it. You get delivery reach in most countries you work in. Plus one contract covering migration, quality, matching, replication, catalog and MDM. Their domain models arrive boxed. Ours do not. You model your domain. We ship no Customer 360 template. Their steward screens have sat in front of plant staff for years. Ours are plainer.
Here is the push-back. It's about what happens after the cutover. The migration ends. The platform bill does not. A lot of their MDM footprints are really migration footprints nobody cancelled. They run two or three domains at enterprise prices. The services channel now belongs to an integrator. Not a software company. So ask two questions. What's the roadmap when the product is a line item inside an SI practice? And when an agent reads the golden record: who checked it was allowed to? Their 2026 copy talks about an AI-ready base. We think ready means the record must pass a quality gate first. It must also pass a consent check. It also means the agent gets refused when it should be.
Questions worth asking both of us
These are the questions we'd want answered as the buyer. Ask them on each call, ours included.
Can you cryptographically prove an audit log entry wasn't deleted?
DataShield: yes. Each record commits to the one before it. Checkpoints are signed and chained. Checking tells deletion apart from truncation and from tampering. Try a sample chain at /verify. Syniti: their platform keeps change logs and approval records. We found no public tamper-proof method. Ask them to show one.
What happens to a revoked agent mid-session?
DataShield re-checks access on each checked tool call. Revocation lands on the very next call, not at the next token refresh. Syniti's public docs describe human roles and workflow approvals. We found no agent session handling of any kind. Ask how long a compromised service account keeps reading master data. After you pull its access.
How does GDPR erasure interact with the audit trail?
DataShield crypto-shreds per-subject key material. It issues an ISO 27560 consent receipt. Actor identities in the chain carry HMAC commitments. The proof still checks out once the subject is gone. Consent is a lawful-basis record per subject and purpose. A missing basis blocks the read, not allows it. Syniti has data quality and oversight tooling that can locate and change records. The erasure method isn't spelled out publicly. Neither is what it does to the log. Ask for the mechanism, not the workflow.
We're a Syniti alternative search that is not doing an SAP migration. Does that change the answer?
It changes it a lot. Most of Syniti's value sits in migration method and SAP depth. You'd pay for both without using either. Say the job is golden records for customer, product or supplier data. Then compare on how the model gets built: match method, stewardship, deployment, price. That's a fight we're happy to have. If S/4HANA is on your roadmap in the next 18 months, say so. Call them first.
Can our own engineers run DataShield MDM without a services engagement?
Yes. The first model is not a blank page. Guided onboarding ships end to end. An interview. An ordered choice gate you check call by call. One step builds the whole model. Everything lands as a draft. Nothing is promoted. No matching runs. The paired-gate check still has to pass. Docker images ship for Auth, Ontology, Corpus and Lighthouse. A signed deploy manifest, Guardian checks it. The whole mastering surface is callable over MCP. Match configs, stewardship calls and golden reads are all scriptable. One honest limit: we ship no boxed 360 domain models. You and the agent decide what an entity is.
Does DataShield have SOC 2?
Not yet. It's planned, not attested, and we won't imply otherwise. Auth runs live with a public threat model. A verifier anyone can run backs it. Guardian and Lighthouse have run live since April 2026. Design-partner terms include source escrow. A small vendor isn't a single point of failure. Details on the security page.
- Capgemini closed its acquisition of Syniti on 3 December 2024, adding more than 1,200 data specialists and citing client demand for complex data migrations to SAP S/4HANA. - Capgemini press release, 3 Dec 2024
- Trade coverage framed the deal as enhancing Capgemini's data features for SAP S/4HANA transformation rather than as a software acquisition. - SDxCentral, Dec 2024
- The Syniti Knowledge Platform is made up of six modules: Migrate, Quality, Match, Replicate, Catalog & Governance, and Master Data Management, with "embedded knowledge from 5,000+ successful global migrations." - syniti.com/platform, 13 Sep 2026
- Syniti's MDM solutions page covers customer, product and vendor domains and names Bridgestone, Phillips 66, Peloton, Zurich, Exxon Mobil and Zespri as customers. - syniti.com, 13 Sep 2026
- The Syniti Knowledge Platform is an SAP Endorsed Application listed on the SAP Store. - SAP Store, 13 Sep 2026
- Third-party estimate, not vendor-published: mid-market Syniti deployments with one or two modules at roughly $150K to $350K a year, full stack at $350K to $1.2M. - VendorBenchmark, 13 Sep 2026
Other head-to-heads
DataShield vs Informatica MDM
The multidomain incumbent. What it costs to stay on it.
MDMDataShield vs Profisee
Mid-market MDM on Microsoft rails, versus self-hosted golden records.
MDMDataShield vs Semarchy
Fast time to value. The evidence layer it doesn't ship.
AllEvery comparison
One honest scorecard per vendor.
See it run before you talk to anyone. Let an agent interview you and build a draft model. Trial a match config. Work a stewardship queue. Then break a live audit chain, and watch the verifier name the damage. Demo Center access is free with a work email.
Get free Demo Center accessYou've seen the proof
Ready for a number? Scope your deployment and we'll price it against your own economics.
Get your quote →