Data reconciliation software for multi-location operator data
Resolve conflicting facts across POS, GBP, Yext, HR, and CMS into one canonical record per location — with source-priority, freshness-tiebreak, and human-routing fallback that learns.
The problem
Your POS says the Denver store is open until 10pm. GBP says 9pm. The website says "See store for hours." Yext says 10pm Monday through Saturday, 9pm Sunday. Which one is right? Whoever logged in last.
Your ops analyst spends Wednesdays going line-by-line through a Google Sheet trying to figure out which source to trust for which field for which location. Multiply that by manager bios that drift between HR and the location page, services that exist in the POS but not the CMS, license status changes that legal forgot to push to marketing, and acquired-brand portfolio merges where four source systems disagree across six brands.
Generic enterprise data-quality platforms like Informatica DQ and IBM InfoSphere solve this for data engineers at $50,000-$500,000 per year with three-to-twelve-month implementations. CDPs like Tealium and Segment resolve customer identity, not location records. SQL deduplication scripts produce CSVs that nobody reviews on time. The operations analyst owns the spreadsheet and the spreadsheet owns the operator.
What success looks like
Every source-system conflict resolves automatically against a policy you configure. POS owns hours. GBP owns categories. HR owns manager bios. The compliance officer owns license status. When sources disagree on the same field, source-priority resolves first. When source-priority is silent, freshness-tiebreak wins. When both are silent, the conflict routes to a single human review queue with full context plus prior decisions on similar conflicts.
Every human decision feeds back as training signal. After 60-90 days, repeat conflicts auto-resolve and the human queue drops significantly. License-status conflicts and HIPAA fields always route through compliance regardless of freshness. Cross-brand portfolio reconciliation handles PE roll-up complexity without manual mapping per acquisition.
The audit trail builds itself. Every reconciliation decision logs source inputs, chosen value, applied policy, and reviewer for six-to-seven-year regulator-defense retention.
How most operators solve this today
Two-tier incumbent stack plus DIY. The middle tier (marketing-specialized) is structurally absent — reconciliation is a data-engineering concept that operators usually delegate to IT.
Enterprise data-quality (Informatica DQ, IBM InfoSphere, Talend Data Fabric, Trifacta/Alteryx)
$5k-$500k/year
Built for data-engineering teams at enterprise scale. 3-12 month implementation through consulting. Not optimized for multi-location operator data shapes (addresses, hours, services, manager bios, license status).
Customer-360 / CDP (Tealium, Segment, mParticle)
$120/mo-$50k+/year
Resolve CUSTOMER identity across touchpoints. Not built for LOCATION-record reconciliation. Different data shapes, different policies, different audit-trail requirements.
SQL deduplication scripts + manual CSV review
Dev FTE + ops analyst time
Data team writes queries; CSV lands in someone's inbox; nobody reviews it on time. No policy engine, no learning loop, no audit trail. Falls apart past 50 locations.
Manual reconciliation FTE
$60k-$90k FTE × 10-20 hrs/wk
Operations analyst owning the master spreadsheet. Workable under 50 locations; becomes the rate-limiter past that.
What changes when this is an agent skill
The Completions reconciliation skill applies a configurable policy stack to every cross-source conflict.
Source-priority resolves first — operator configures which system wins for which field per vertical. Freshness-tiebreak resolves the rest, with configurable staleness thresholds. Anything still unresolved routes to a single human review queue with full conflict context plus the reviewer's prior decisions on similar conflicts. License-status fields and HIPAA-flagged data always route through compliance regardless of policy.
The learning loop sharpens over time. Every human decision feeds back as training signal — after 60-90 days, the auto-resolution rate climbs and the human queue shrinks to genuine edge cases. Per-vertical schema awareness applies stricter rules where regulators require them. Per-brand-id configuration handles PE roll-up portfolios sharing back-office systems across acquired brands.
Audit trail is built in. Every decision logs source inputs (system + field value + timestamp), chosen canonical value, applied policy (source-priority / freshness / human-routed), reviewer when human-routed, and downstream emission outcome. Composes with the versioned-history skill for six-to-seven-year regulator-defense retention.
Operator-friendly cost ($2,000-$4,000/mo early-adopter) replaces $50,000-$500,000/year enterprise data-quality stacks plus the operations-analyst FTE labor.
Agents that include this skill
Skills live inside agent rentals. To get this skill in production, hire any of the agents below — context-tuning at onboarding is included in the first month.
Master Record Canonicalization Agent
Ingests every operator source system, resolves per-location fact conflicts, and emits the canonical master record downstream agents consume.
Early-adopter
$2,000–$4,000/mo
FAQ
- What is data reconciliation software?
- Software that resolves conflicting facts across multiple source systems into a single canonical record. The Completions skill applies source-priority plus freshness-tiebreak plus human-routing fallback specifically to multi-location operational data.
- How is this different from Informatica or IBM InfoSphere?
- Those tools are enterprise data-quality platforms aimed at data-engineering teams. Implementation runs 3-12 months. This skill is purpose-built for multi-location operator data (addresses, hours, services, manager bios, license status) with operator-friendly policy configuration.
- Is this the same as entity resolution?
- Functionally yes — the data-engineering term for the same operation is entity resolution. The Completions framing uses operations-buyer language because the primary buyer is the operations director, not the data engineer.
- How does the learning loop work?
- Every human review feeds back as a training pattern. Over 60-90 days, the proportion of conflicts auto-resolved by policy climbs significantly. The human queue shrinks to genuine edge cases the policy stack cannot reach.
- What if multiple sources are equally trustworthy for the same field?
- Configure them as co-equal sources; freshness-tiebreak resolves. If freshness is equal too, the conflict routes to a human reviewer with full prior-decision context.
- What about regulator-required fields like license status or HIPAA-flagged data?
- Per-vertical schema awareness applies stricter rules. License-status conflicts always route to the compliance officer regardless of freshness. HIPAA fields require explicit human confirmation per record.
- How does this compose with multi-source ingestion?
- The multi-source-ingestion skill pulls per-source data; this skill resolves cross-source conflicts before the canonical record is promoted. Both feed the master-record-sync skill that emits change events to downstream marketing systems.
- What is the audit trail format?
- Every decision logs source inputs (system + field value + timestamp), chosen value, applied policy (source-priority / freshness / human-routed), reviewer if human-routed, and downstream emission outcome. Composes with the versioned-history skill for 6-7 year retention.