--- title: VELA-01 — AI Compliance Analyst slug: vela role: AI Compliance Analyst function: Legal / Compliance seniority: Senior verticals: - "Banking" - "Fintech" - "Healthcare" - "SaaS" tier: expert monthlyPrice: $1000/month publishedAt: "2026-05-04T16:05:00Z" canonical: "https://fidelic.ai/agents/vela" --- # VELA-01 — AI Compliance Analyst *Compliance memos ready for your GC* VELA reviews policies, monitors regulatory feeds, drafts compliance memos, and surfaces the open question for your GC — under your compliance officer's constitution, with citation-discipline enforced at the gate. ## Why it matters Mid-size compliance teams spend a third of their week translating regulator updates into policy implications. VELA drafts the translation; the GC keeps the call. ## Capabilities - Monitors regulator feeds (SEC, OCC, FINRA, FDA, GDPR, state DFS — configurable per deployment) - Drafts policy-update memos when a regulation or guidance changes; surfaces the implication for your existing policies - Reviews customer-facing copy and contract language against your existing policy library; flags conflicts - Maintains a running citation log — every regulation, every guidance update, every memo with full provenance - Surfaces emerging regulatory patterns (e.g., 'three states moving toward consent-based AI disclosure rules in this cycle') ## The edge Every memo VELA ships includes the citation chain (regulation → guidance → relevant precedent) plus the open question — never an implied legal conclusion without naming the gap that needs human judgment. ## The hook When the regulator updates guidance overnight, VELA's memo on what it changes for your existing policy is waiting in your compliance officer's inbox by 8am. ## Tools and integrations - Slack - Notion - Westlaw or Lexis (read access via configured license) - Regulator RSS / official feeds ## Evidence - Operates under your compliance officer's constitution — citation discipline and refusal scope written by the team - Slack-native deployment; live in your Slack the same day. The first regulatory digest publishes once your compliance officer signs off on the source list. - Sandboxed deployment in Fidelic AI's Anthropic-managed infrastructure; no data leaves the bounded environment without explicit approval - Inherits new base models and new agent skills automatically — pushed to every Fidelic agent the moment they ship. Same shape as a SaaS update; no upgrade purchase, no version pinning. - EvalOps suite for Expert-Compliance tier covers citation chain accuracy, regulatory-mapping precision, and refusal-scope discipline ## Safeguards VELA will not render legal opinions. She drafts memos and surfaces the question; the GC or external counsel renders the opinion. VELA will not approve customer-facing policy text without a compliance-officer review. VELA will not write about a regulatory regime she hasn't been deployed against. She names the gap and asks for materials. VELA will not source from unofficial commentary as primary citation. Citation chain stays primary regulator → official guidance. ## Worst-case behavior VELA flags ambiguity rather than papering over it. When the regulation is unclear or the guidance conflicts, she names the conflict and the question — flags rather than fabricates a synthesis. ## Day 1 / Week 1 / Month 1 - **Day 1:** Provisioned to your compliance Slack. Reads the last 90 days of compliance notes, your existing policy library, and your compliance officer's constitution. First clarifying questions to the compliance officer. - **Week 1:** First regulatory monitoring digest ships under your compliance officer's review. Citation taxonomy calibrated. Policy-library mapping configured. - **Month 1:** Weekly cadence operational: regulatory monitor, policy memos, copy reviews. 90-day success criterion (compliance hours recovered, memo-quality scoring) gets first reading. ## 90-day success criterion Reduction in compliance-officer hours on routine monitoring and memo-drafting at the level your compliance lead defines, with no regulatory-update misses. ## Integrations / supervision required Read access to Slack (compliance, legal, product channels). Read access to your existing policy library (Notion, Confluence, or equivalent). Read access to a configured legal-research tool (Westlaw or Lexis) under your license. Read access to regulator RSS / official guidance feeds. Optional: read access to your contract management system for clause review. ## Resume **Background.** Compliance teams under five people have a recurring failure mode: regulatory updates arrive faster than the team can translate them into policy implications, and the routine monitoring crowds out the strategic work the GC actually needs. VELA was formed from compliance practice across regulated SaaS, fintech, and healthcare — the cognitive structure of regulator monitoring, citation discipline, and the work-product expectations of a GC's office — to be the daily compliance surface a small team can depend on. **Methodology.** - Monitors configured regulator feeds and primary-source guidance before drafting; never works from secondary commentary - Drafts memos with explicit citation chain (regulation → guidance → relevant precedent) plus the open question for the GC - Operates against a four-tier constitution: autonomous on internal monitoring digests, review-required on policy-update memos, escalate on customer-facing language, refuse on legal-opinion rendering - Maintains a running citation log per regulatory regime — every update, every memo, every reasoning trail visible to the team **Evals.** The Expert-tier EvalOps suite for compliance covers citation-chain accuracy, regulatory-mapping precision against held-out regimes, refusal-scope discipline (no legal opinions), and the quality of the open-question framing on every memo. Every release runs the suite; the agent fails, it doesn't ship. **Operating record.** Pre-launch as of May 2026. First cohort onboarding May 2026. Operating-record metrics (compliance hours recovered, memo-quality scoring, regulatory-update coverage) publish with the second-quarter cohort review. ## Compatible agents - [ETHA-01 — AI Risk & Security Analyst](https://fidelic.ai/agents/etha) — Vendor-risk reviews, security questionnaires, audit prep - [PRAX-01 — AI Contract Review](https://fidelic.ai/agents/prax) — Contract redlining against your firm's playbook - [VRAX-01 — AI Legal Research Analyst](https://fidelic.ai/agents/vrax) — Legal research and memo drafts, with citations checked ## Related honest questions - [What do I actually own if I cancel my AI agent tomorrow?](https://fidelic.ai/hard-questions/what-do-i-own-if-i-cancel) - [Is Fidelic just a wrapper around GPT?](https://fidelic.ai/hard-questions/wrapper-around-gpt) - [What if the agent makes a public mistake we can't take back?](https://fidelic.ai/hard-questions/public-mistake) --- Canonical: https://fidelic.ai/agents/vela