Fairmarkit

Overview: Boston-based AI sourcing platform, founded in 2017 (co-founder and CEO Kevin Frechette). Fairmarkit started in tail-spend RFQ automation and now positions itself as “autonomous sourcing for all spend.” On April 29, 2026 it launched Total Agentic Sourcing: a network of AI agents, led by the KIT assistant, that sources both tail and strategic spend (from a $500 purchase to a $500M contract) in one environment. It sits on top of SAP Ariba, S/4HANA, Coupa, Oracle and ServiceNow rather than replacing them. It has raised about $78M in total. The last round was a $35.6M Series C in September 2022, led by OMERS Growth Equity, with GGV, Insight Partners and ServiceNow (strategic) taking part. Customers include Boeing, BP, Walgreens, Goodyear, Snowflake, Sonoco and Emirates Flight Catering.

Core Platform & Architecture

1. Deployment models Cloud SaaS only. Deployed as a sourcing layer on top of an existing ERP/P2P stack. No on-premise option has been disclosed.

2. Data model — unified or modular/siloed A sourcing-focused layer, not a full S2P platform. Its own data model covers intake requests, RFx events, supplier bids, awards and savings/cycle-time metrics. Purchase requisitions, POs, invoices and payments stay in the connected ERP/P2P system, and awards are written back to it through bi-directional sync.

3. Technology stack Cloud-native platform with a multi-agent AI architecture: the KIT assistant coordinates specialised agents. Supplier matching draws on historical bid and response data across its network. LLM providers and infrastructure are not publicly disclosed.

4. Uptime/SLA guarantees Fairmarkit says it is SOC 2 Type 2 certified. Uptime/SLA commitments are not publicly disclosed and are presumed to be contract-defined.

AI & Innovation

5. AI/ML capabilities — native vs bolted on Native. Fairmarkit has used ML for supplier recommendation and bid analysis since its tail-spend origins, and the 2026 platform is organised around AI agents.

6. Generative AI support Yes. The Intake Agent gathers requirements through a guided conversation. The RFx Agent builds event content automatically, and KIT gives buyers a conversational interface. The 2026 launch included a partnership with negotiation author Chris Voss.

7. Agentic/autonomous workflow layer Yes, and this is the core of the product. Total Agentic Sourcing uses an Intake Agent, a Supplier (Discovery) Agent, an RFx (Execution) Agent, an Evaluation Agent and a Performance & Compliance Agent. Low-value requests can run end to end without a buyer.

8. AI in supplier/vendor evaluation and scoring The Supplier Agent matches suppliers from a network of 2.7M+ using category fit and a score for how likely each supplier is to respond. It also tracks supplier diversity (MWBE, WOSB, SDVOSB, LGBTBE). The Evaluation Agent compares bids, models scenarios, calculates TCO and suggests where to negotiate.

9. Predictive analytics Limited to sourcing: it predicts which suppliers will respond, recommends the best supplier and tracks savings and cycle times. It offers no broader spend forecasting or supply-risk prediction.

10. AI feature update frequency and public roadmap Fairmarkit ships AI features quickly, moving from ML-driven autonomous sourcing to Total Agentic Sourcing in April 2026. It has no public roadmap, and the release cadence is not disclosed.

Integrations & Interoperability

11. ERP and financial system integrations This is Fairmarkit’s strongest area. It has bi-directional API integrations with SAP S/4HANA, SAP ECC, SAP Ariba, Coupa, Oracle, ServiceNow and Zip. It picks up purchase requests in those systems, runs competitive sourcing, and writes the winning supplier and price back to the PR.

12. Open API Integrations are API-based. A public help centre exists (docs.fairmarkit.com), but public developer API documentation is limited.

13. SSO and identity/auth standards Enterprise SSO is supported. Companies can also require suppliers to log in through the customer’s SSO before they see full RFQ details. Fairmarkit does not publicly detail which SAML or OIDC standards it supports.

14. Pre-built connectors to supplier networks or third-party data providers Its own network holds 2.7M+ suppliers. Amazon Business is connected for quotes (500K+ sellers). Other connectors include Candex (supplier onboarding and payment for one-off suppliers), Supplier.io (diversity data) and SpendHQ (turning spend analysis into sourcing events). 2025 partners: Conduent, Inception (G42).

Tender & Procurement-Specific Functionality

15. RFx types supported Supported event types are RFQ (its core), RFI, RFP and complex strategic events, plus auctions, which were added with Total Agentic Sourcing. The RFx Agent can bundle similar RFQs into one event. It has not been shown to match Coupa, Ariba or Jaggaer on advanced optimisation.

16. Compliance and audit trail Auto Compliance applies rules by category, region and spend threshold and records an audit trail for each event. Supplier-diversity reporting is included, and Fairmarkit says it is SOC 2 Type 2 certified. Fairmarkit does not advertise support for public-sector tender rules.

17. Configurability — custom workflows, scoring criteria, approval hierarchies Buyers can configure sourcing rules, thresholds, custom fields and evaluation criteria. Approval hierarchies and wider workflow routing usually stay in the host ERP/P2P or intake tool (such as Zip or ServiceNow).

Product Maturity & Roadmap

18. Known gaps and limitations

Not an I2P/S2P suite. Sourcing (intake-to-award) is native, and intake is native for sourcing requests only. Contracts/CLM, POs, invoices and payments are not native and depend on the ERP/P2P system. Supplier management is limited to the supplier portal, matching and diversity data.
Strategic sourcing and auctions were only extended in April 2026, so their maturity against established strategic sourcing suites is unproven. Its strength is still tactical and tail spend.
Depends on an ERP/P2P system, so it offers limited standalone value to organisations without an existing procurement stack.
G2 and Capterra reviewers flag reporting that is too basic and hard to customise, analytics columns that can’t be edited, poor tracking of unread supplier chat messages, few unit-of-measure options and supplier recommendations that are sometimes inaccurate. G2 users also cite integration issues and extra costs.
No new funding since 2022 (about $78M in total), which is modest capital compared with suite vendors and well-funded agentic rivals.
Pricing, SLA terms and developer API documentation are not public.
Not built for public-sector tendering compliance or for direct-materials and supply-chain processes.