ooligo

Kombo

unified-api ats-integrations · hris-integrations · payroll-integrations · embedded-integrations
API
Recruiting & TA
7.6 /10

What it is

Kombo is a common-model API for HR data, built for software vendors rather than for the HR team itself. A recruiting-tech or HR-tech product integrates once against Kombo’s common data model, and that product’s customers connect their own ATS, HRIS, payroll, LMS or assessment system through Kombo Connect, an embeddable connection flow. Kombo runs four category APIs (ATS, HRIS and payroll, assessment and background check, LMS) across what it calls “hundreds” of systems; competitor write-ups put the total at 250+ as of October 2026, and Kombo’s own ATS write-back post covers 108 ATS integrations.

Kombo was founded in 2022, has offices in Berlin and New York, and counts Y Combinator among its investors. On 18 February 2026 it raised a $25M Series A led by Volition Capital, bringing total funding to $30M. It says it works with 300+ companies and names Indeed, Paychex, HireVue, Mercor, Juicebox and Gem as customers; Greenhouse, Metaview, SeekOut and TestGorilla appear on its logo wall. It reports SOC 2 Type 2 and ISO 27001, and hosts in the US, Canada or the EU.

Why it shows up in recruiting stacks

Ops leaders rarely buy Kombo directly. It shows up because a tool you are evaluating (Gem, Juicebox, HireVue and many smaller AI recruiting products) uses it to reach your ATS. That matters in two seats. If you run integrations at a recruiting-tech vendor or an internal TA engineering group, Kombo is the buy option in “build our own ATS connectors, use Merge, or use Kombo?”. If you are buying one of those tools, Kombo is the sub-processor holding a copy of your candidate data, and it belongs in your security review.

The HR-only focus is the reason to pick it. Write-back depth is where common-model APIs usually thin out, and Kombo publishes its numbers: per its October 2026 ATS post, it creates applications on 103 of 108 ATS integrations, moves candidates between stages on 52 and rejects with a reason on 24. Those are the actions an AI screening or sourcing product needs to close the loop inside the customer’s ATS instead of a side tab. Kombo also claims certified partnerships with Workday, SAP SuccessFactors and Oracle, the systems where partner access, not code, is the bottleneck.

Pricing reality

Kombo publishes no prices. All three plans (Start, Scale, Enterprise) are sold through a demo. The model is a fixed annual platform fee plus a flat fee per connected end customer that falls with volume, with no per-call or data-volume charges; inactive integrations are not billed. A free proof of concept with sandbox access is offered.

What the tiers gate matters more than the missing price list:

  • Start is read-only. There is no write API on this plan.
  • Scale adds read/write access, faster syncs, enterprise systems such as Workday and iCIMS, 30-day logs and a dedicated account manager. Any product that writes back into an ATS needs Scale at minimum.
  • Enterprise adds multi-region data residency, SAML SSO, uptime and support SLAs, 90+ day logs, SFTP and custom integrations, and SAP and Workday sandboxes.

The only public spend figure is Cledara’s marketplace average of about $5,200 among its customers, which states neither the period nor the number of connections; read it as a small-deployment signal, not a quote. For an anchor, Merge publishes its common-model API Launch plan at $650 a month for up to 10 linked accounts plus $65 for each one after, which is $15,600 a year at 20 connected customers. Ask Kombo to price the same 20- and 100-customer scenarios on Scale so the two quotes are comparable.

Best for

The head of integrations or platform engineering at a recruiting- or HR-tech vendor, Series A through growth stage, whose customers sit on dozens of ATSs and HRISs including Workday, iCIMS and SuccessFactors, and whose product must write candidates, stage changes or results back into those systems. An EU-heavy customer base strengthens the case, given EU hosting and a Berlin team.

Not for

  • An in-house TA ops team connecting its one ATS to its one HRIS. Use the native integration or an iPaaS; a common-model API pays off only when you serve many customers on many systems.
  • A product that also needs CRM, accounting or ticketing integrations. Kombo is HR-only.
  • An agent team that needs MCP access to live HR data today. Kombo’s September 2026 blog says its integrations can be used via MCP, but a competitor review found only a documentation MCP server in Kombo’s public docs as of October 2026. Confirm it in the demo before you plan around it.

Versus the alternatives

  • Merge is the largest common-model API vendor and covers HRIS, ATS, CRM, accounting, ticketing and file storage, with published self-serve pricing. Pick Merge when you need categories beyond HR or want to start without a sales call.
  • Build in-house when most of your customers sit on a handful of ATSs with open APIs, such as Greenhouse, Lever and Ashby. You keep the data path and the roadmap, and you pay in engineering time for every new system and every API change.
  • Knit and Unified.to are the newer challengers. Both sell pass-through reads instead of a stored copy, and both pitch MCP access for agents. Pick one when your customers’ security reviews reject a third party storing candidate PII, or when your agents need reads fresher than a sync cycle.

If none fit, build the three or four connectors your largest customers need and add a common-model API only when the long tail of ATS requests starts to block deals.

Watch-outs

  • Kombo stores a copy of the data. It mirrors candidate and employee records into its own database, syncs every 3 hours by default (faster on paid tiers, down to five minutes per Kombo), and receives source webhooks on only 22 integrations. Guard: list Kombo as a sub-processor in your DPA, choose the hosting region at setup, and trigger a manual sync before any time-sensitive action such as auto-rejecting a candidate.
  • Write coverage is uneven. Stage moves work on 52 of 108 ATSs and rejection with a reason on 24. Guard: before signing, build a matrix of your top 20 customer ATSs against the exact writes your product needs and test each one in the proof-of-concept sandbox.
  • Enterprise sandboxes are slow. G2 reviewers, as summarized by Knit, report waiting three to six months for an iCIMS sandbox. Guard: file partner sandbox requests on day one, and if you sell into Workday or SAP shops, budget for Enterprise, where those sandboxes are included.
  • Default credentials create switching cost. Customers connected through Kombo’s default credentials must reconnect if you move to another provider, according to a competitor review of Kombo’s docs. Guard: use your own partner credentials wherever the ATS supports them, and write token and data export into the contract.