Vendor risk tiering

Use case Β· Compliance

← All use cases
β—£

Vendor risk tiering

Compliance

Given a vendor and the data they touch, assign a risk tier and the due-diligence depth required β€” so you don’t pentest your snack supplier.

The routing recipe
Job typeClassification / routingPolicyBalanced
routes to
GPT-OSS 20B groq
groq:openai/gpt-oss-20b
90.9
match score
84.6
task fit
100
cost
96.4
speed
Why this model from benchmarks & capabilities
  • "Classification / routing" leans hardest on instruction; GPT-OSS 20B scores 78/100 there β€” #23 of 40.
  • Strength on this class of work shows up on IFEval, MMLU (zero-shot), where OpenAI (open-weight) models are competitive.
  • Cost: free tier. Speed: ~1000 tok/s on Groq LPU.
  • Chosen over GPT-OSS 120B for cheaper & faster.
Capability the job needsWeightGPT-OSS 20BCatalog rank
instruction0.9078#23 of 40
speed0.9096#1 of 40
knowledge0.3070#26 of 40
Relevant benchmarksIFEval MMLU (zero-shot) EconomicsCost: free tier.
Speed: ~1000 tok/s on Groq LPU.
vs runner-upChosen over GPT-OSS 120B for cheaper & faster. (GPT-OSS 120B, score 83.9)
Example result
Sample input β–Ύ
Vendor: a transactional email provider that processes customer email addresses and sends on our behalf; has API access to our app.
Tier: High. Why: processes customer personal data (emails) and holds standing API access to your app β€” a compromise reaches your users. Due diligence: SOC 2 Type II review (or ISO 27001), signed DPA with SCCs, a security questionnaire (auth, encryption, breach SLA, sub-processors), and scoped API keys with least privilege. Reassess annually or on a security incident.
Test it on your own data
Sign in to continue

LLM Switchboard is private β€” sign in with Authlee to access the control room.

Sign in with Authlee
← Back to home