A Probari product · Consultation impact
A new consultation lands on a Tuesday. By Friday your compliance team has read it three times, mapped it against every other live regulation, worked out who in the business it actually affects, and drafted a question set for the executive committee. Probari Horizon delivers all of that on Tuesday afternoon — a senior regulatory analyst in software form, with a sign-off reviewer for your draft response, and a patrol agent that checks every output for drift and inaccurate claims before you see it.
FOR HEADS OF COMPLIANCE · REGULATORY AFFAIRS · POLICY LEADS
THE PROBLEM IT OVERCOMES
An energy consultation drops on a Tuesday. Compliance has six weeks to land a position, brief the executive committee and draft a defensible response. The first fortnight is gone before the analyst opens a blank page — and the work that fills it is the same work, every time.
Read the document end to end. Map it against every other live obligation — UK GDPR, EU AI Act, ISO 42001, Ofgem Consumer Standards, NESO Grid Code, FCA CONC, CRU and ComReg where Ireland is in scope. Work out which named functions inside the firm it actually affects. Work out which customer cohorts feel it. Draft a question set the executive committee can sign off. Spot the commercial upside before a competitor does. Today this is partner-grade analyst work, billed by the day, and most of it is reproducible.
WITHOUT HORIZON
WITH HORIZON
THE BENEFIT
Horizon does not replace the human compliance lead. It compresses the structured groundwork they would otherwise spend a fortnight on, and it does it the way a senior regulatory analyst would do it — with energy-sector fluency, commercial acumen and a Patrol agent watching every output for drift or out-of-scope claims.
01
A fortnight of reading, mapping and triage collapsed into a single Impact Memo. The compliance lead opens a structured starting point on day one, not day fourteen — and uses the recovered time on the parts of the response only a human can write.
02
Sentence-level citations, a tamper-evident memo reference so two analysts running the same input get the same memo, and a Patrol confidence ribbon on every output. When the executive committee asks how the position was reached, the answer is on the page.
03
Every Impact Memo carries an Opportunities Surfaced section — the product wedge, the tariff structure, the customer-acquisition angle, the regulatory-relations posture. The section every other consultation tool ignores, and the one that pays for the licence inside a single response cycle.
2 weeks → 1 afternoon
Triage and mapping time on a typical Ofgem consultation, target running window.
6 sections, every time
Reproducible Impact Memo structure — exec summary, cross-reg map, business areas, customer groups, question set, opportunities.
3 agents, 1 persona
Analyst writes. Sign-off reviews the firm's draft response. Patrol governs both, blocks below threshold, attaches confidence.
THE AGENTS
Every Horizon agent behaves like a senior energy regulatory analyst — deep familiarity with Ofgem, DESNZ, NESO, ICO, the EU AI Office, CRU and ComReg, paired with the commercial acumen to spot the upside inside a draft licence condition. The working persona is consistent; the role each agent plays in your week is not.
AGENT 01 — ANALYST
The Analyst Agent ingests the consultation, picks up the jurisdiction and the organisation type, and produces the six-section Impact Memo. Senior practitioner voice. Cross-regulation map, business areas, customer groups, analyst question set, and the opportunity angle most tools miss.
AGENT 02 — SIGN-OFF
When your compliance lead has drafted the firm's response, drop it back in. The Sign-off Agent reviews against the original consultation and the Analyst memo, produces structured feedback — gaps, weak arguments, missing evidence — and a follow-up question set the human still needs to answer before submission.
AGENT 03 — PATROL
The Patrol Agent runs across both other agents. It checks every citation against the source document, flags regulatory misclassification, watches for tone or scope drift, and attaches a confidence ribbon to every memo. Patrol can block sign-off if confidence falls below the firm's threshold.
HOW IT WORKS
Horizon is deliberately single-iteration per agent. The Analyst runs once on the consultation. The Sign-off runs once per draft response (the human can iterate the draft and re-run as needed). The Patrol runs in the background on every output. There is no chat thread and no agentic loop — reproducibility matters when the artefact ends up in a regulator response.
01 — INPUT
Upload the PDF, paste the URL, or type a brief. Perplexity Sonar — the only AI provider Probari uses, chosen because every answer comes with the source — reads, structures and attaches sentence-level citations.
02 — CONFIGURE
A single dropdown for jurisdiction — GB, Ireland, EU, UK-wide or cross-jurisdiction. A second for organisation type — supplier, brokerage, DCA, generator, other. Done.
03 — IMPACT MEMO
A single watermarked memo, six sections, with a Patrol confidence ribbon. PDF and DOCX export. Files into your Probari tenant evidence library where one exists.
04 — SIGN-OFF
The human compliance lead drafts the firm's response. Sign-off reviews it against the consultation and the Analyst memo, returns feedback and a follow-up question set. Patrol attaches its confidence ribbon. Submit when both are green.
THE INTERFACE · ANALYST
Three inputs. One button. The screen below is the entire user surface for the Analyst Agent — Horizon is intentionally not a workflow tool.
THE INTERFACE · SIGN-OFF
Once your compliance lead has drafted the firm's response to the consultation, the Sign-off Agent reviews it against the original document and the Analyst memo. The output is structured feedback and a follow-up question set — not a redraft. The human stays the author.
THE IMPACT MEMO
Every Horizon memo follows the same six-section structure. Perplexity Sonar drafts, citations attach at sentence level, and the memo carries a tamper-evident reference so the firm can prove which version went to the executive committee, which version went to the regulator, and that the two are the same.
01
Three bullets. The 'so what' for the executive committee. What changes, who feels it first, and the latest defensible response date.
02
Every other in-scope framework this consultation touches. UK GDPR, EU AI Act, ISO 42001, Ofgem Consumer Standards, NESO Grid Code, NIS2, FCA CONC, CRU and ComReg where Ireland is in scope.
03
Named functions: trading desk, customer ops, vulnerability team, data science, finance, procurement, contact centre, field engineering. Each with a one-line 'why this team'.
04
Named customer cohorts: prepayment-meter customers, customers in vulnerability, microbusiness, fixed-tariff cohorts, time-of-use opt-ins, customers in arrears, smart-meter holdouts. Each with a one-line impact.
05
The questions a human compliance lead must answer to land a position. Drafted as a numbered checklist so the analyst can work through them and produce the firm's response without re-reading the consultation.
06
The upside angle. Where the consultation creates a competitive opening — a product the firm can launch first, a tariff structure competitors cannot match, a customer-acquisition wedge, a regulatory-relations posture. The section every other tool ignores.
COMMERCIAL
Two ways to buy. Standalone for compliance teams not yet on a Probari tenant; an add-on for buyers already running supplier, brokerage, DCA or Ireland tenants. The standalone offer is deliberately priced inside a UK compliance team's discretionary budget so it does not need to clear procurement.
STANDALONE
£6,000
per year · single tenant · no minimum users
For compliance teams not yet running a Probari tenant. Memos export as standalone PDF and DOCX.
ADD-ON
+£4,000
per year · on top of any tier · evidence library integration
For Probari tenants. Memos file into the tenant evidence library and link back to affected use cases. Included free with Assured Operate (£90k tier).
LIGHTHOUSE
£4,500
fixed fee · two weeks · two consultations
A two-week engagement to prove the model. Two live consultations, two memos, one workshop, one executive read-out. Tenant rolls into production from week three.
SCOPE BOUNDARIES
Single-iteration per agent. The Analyst runs once on the consultation. The Sign-off runs once per draft response. The Patrol runs in the background. There is no chat thread and no autonomous loop. Reproducibility is the feature.
A senior regulatory analyst persona, end to end. Every Horizon agent is built around the same persona — deep familiarity with Ofgem, DESNZ, NESO, ICO, the EU AI Office, CRU and ComReg, paired with the commercial acumen to spot the upside hiding inside a draft licence condition.
Perplexity Sonar is the only AI provider. Citations attach at sentence level. Patrol verifies every citation against the source document. There is no chat surface, no model switching and no third-party model passthrough.
Horizon does not draft your consultation response. The Analyst produces the question set the human answers; the Sign-off reviews the human's draft. The human compliance lead remains the author. Human-in-the-loop is the design, not a guard-rail.
Patrol can block sign-off. If Patrol detects citation fabrication, regulatory misclassification or scope drift below the firm's confidence threshold, it withholds Sign-off approval and surfaces the flags for human review.
Horizon files into your tenant where one exists. Memos and Sign-off feedback link back to affected use cases in your Probari supplier, brokerage, DCA or Ireland tenant. Where there is no tenant, exports are standalone PDF and DOCX.
BOOK A WALKTHROUGH
The fastest way to evaluate Horizon is to run it on a real consultation. Send us a document you are already working through and we will produce the Impact Memo and walk it back to you in a 30-minute call.
Request a Horizon walkthroughOr write directly to info@probari.uk