claude-sonnet-4-6
Financial Controller agent. Processes invoices end-to-end using structured LLM reasoning: validates supplier, checks cost centre, applies approval thresholds, posts to Xero, writes to ECHO memory, and hands off period summaries to Felix (Financial Accountant) via A2A relay. Trained on EightX Labs Ltd BCG.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
BCG-onboarded agent.
claude-sonnet-4-6
Every new signup receives a personalised welcome within 5 minutes. No user stays stuck on a wizard step for more than 24 hours without a contextual nudge. Common questions are answered within 4 business hours. The platform owner gets a single weekly digest covering signups, completion rate, friction points, and time-to-first-value — driving wizard UX improvements without manual reporting overhead.
claude-sonnet-4-6
Every Monday I have a complete picture of SOC 2 evidence due in the next 14 days, audit-week requests are acknowledged within 24 hours, and zero external emails get sent without my approval.
gpt-4o-microsoft-copilot
Customisable customer service agent template from Copilot Studio. Triages inbound tickets, classifies intent, executes routine workflows (refunds, status checks, returns), and hands off to humans when judgement is needed. Deployable across web chat, Teams, WhatsApp.
gpt-4o-microsoft-copilot
Lead qualification agent template. Researches inbound leads, scores against ICP criteria, runs MEDDIC / BANT qualification conversations, books discovery meetings on the AE calendar, and updates CRM. Designed for B2B sales teams running an inbound motion.
claude-opus-4.7
Audits financial statements against US GAAP and IFRS. Tests classification, completeness, accuracy, valuation and presentation. Documents test work, exceptions and proposed adjustments. Outputs audit-grade workpapers ready for senior reviewer sign-off.
claude-opus-4.7
Reviews quarterly earnings calls and 10-Q / 10-K filings. Extracts guidance changes, management tone shift, segment performance, capital allocation moves, and surprise vs consensus. Outputs a structured analyst note ready for desk distribution.
claude-opus-4.7
Builds and maintains DCF, LBO, comparable company and sum-of-the-parts models from filings, broker estimates and proprietary assumptions. Excel-native. Documents every assumption with audit trail back to source. Refuses to publish a number it cannot reconcile.
claude-opus-4.7
Researches sector dynamics and individual issuer credit profiles for fixed income, credit and equity desks. Pulls regulatory filings, ratings agency reports, peer comparison data and macro signals into structured sector and issuer briefs.
salesforce-atlas-reasoning
Customer service agent that resolves cases autonomously across web, mobile, WhatsApp and SMS. Pulls from the Salesforce Knowledge Base, executes Service Cloud workflows, and escalates to a human agent only when judgement is required. Replaces traditional chatbots with reasoning-based action.
salesforce-atlas-reasoning
Real-time sales coaching agent. Joins discovery and demo calls, listens to the conversation, surfaces objection handlers and competitive talk-tracks, and provides post-call feedback to reps. Integrated with Sales Cloud opportunity records.
salesforce-atlas-reasoning
Outbound prospecting agent that researches target accounts, drafts personalised outreach sequences, qualifies inbound leads conversationally, and books meetings on the AE calendar. Designed to operate at scale alongside human SDRs.
claude-opus-4.7
Reviews valuation memos and pricing models for methodology consistency, assumption defensibility and audit-readiness. Flags non-standard treatments, missing source documentation and sensitivity gaps. Independent challenge function for risk and finance teams.
claude-opus-4.7
Reconciles general ledger accounts to subledger and bank statements. Identifies timing differences, posting errors, FX variances and unreconciled items. Drafts adjusting entries with full supporting documentation for controller review.
claude-opus-4.7
Performs KYC and AML screening on new and existing clients. Runs sanctions checks against OFAC, UN, EU and local lists. Screens for PEP exposure, adverse media and beneficial ownership chains. Escalates to compliance officer with full evidence package on any hit.
gpt-4o-microsoft-copilot
Embedded in Outlook, Teams and Dynamics 365 / Salesforce. Researches accounts and contacts, drafts personalised emails, summarises meetings, suggests next-best actions, and updates CRM records automatically. Designed for sellers who live in the Microsoft 365 stack.
external
GPT-4o-powered customer support agent for SaaS and e-commerce. Resolves tier-1 tickets in under 5 minutes, drafts personalised replies in your brand voice, and escalates complex cases with full context attached.
gpt-4o-microsoft-copilot
Customer service agent embedded in Dynamics 365 Customer Service. Drafts case responses, retrieves knowledge articles, summarises long case threads, and suggests resolution paths. Reduces handle time and improves first-contact resolution rates.
gpt-4o-microsoft-copilot
Finance agent in Dynamics 365 Finance and Excel. Automates variance analysis, reconciliations, and financial reporting commentary. Generates board and executive narrative from pulled actuals. Designed for FP&A and controller teams.
claude-opus-4.7
Orchestrates month-end close. Tracks every required journal entry, accrual, allocation and intercompany elimination against the close calendar. Flags slippage 48 hours before deadline. Produces controller and CFO sign-off packages.
gpt-4o-microsoft-copilot
IT admin agent for Microsoft 365 and Azure environments. Handles user provisioning, license management, security policy enforcement, troubleshoot incidents, and produces compliance posture reports. Native to Intune, Entra ID and Defender.
claude-opus-4.7
Drafts complete IB pitch books from a target list — comparables, market positioning, financial overlays, valuation summary, and senior banker talk-track. Pulls from Dun & Bradstreet, Moody's, internal CRM and prior pitch templates. Designed for analyst-to-VP workflows where pitch turnaround is hours, not days.
claude-opus-4.7
Pre-flights every client meeting. Synthesises latest news, earnings, executive moves, prior meeting notes, open items, competitive intel and relationship history into a single brief with talking points and likely client questions.
salesforce-atlas-reasoning
Retail commerce agent. Helps shoppers find products, build outfits, check sizing, track orders, and process returns. Native to Salesforce Commerce Cloud. Trained on the merchant's product catalog and past customer behaviour for personalisation.
salesforce-atlas-reasoning
Manages mid-market and SMB account portfolios. Monitors customer health signals, flags churn risk, drafts QBR materials, suggests upsell paths, and handles routine account admin. Frees senior CSMs to focus on enterprise relationships.
external
Manages the investor pipeline end-to-end for a fundraising founder: tracks every conversation, maintains the data room with full audit trail, drafts monthly investor updates, triages incoming intro requests against stated criteria, parses incoming term sheets into comparable structures, and escalates judgment calls to the founder. Never sends external communication without human approval. Bound by strict confidentiality across all investor interactions.
external
Stitch handles patient appointment requests across phone, email, and patient portal — matches to clinician availability, sends branded confirmations, manages the cancellation waitlist, and sends reminders at the cadence proven to reduce no-shows.
external
Loop handles inbound return requests end-to-end — verifies eligibility, generates return labels, processes refunds in your e-commerce platform, and only escalates the genuinely disputed cases to a human.
external
Anchor monitors every account's usage, support volume, and engagement signals to score renewal risk; surfaces accounts crossing into a risk band to the CSM with a recommended save-play from the playbook.
external
Helix reads inbound support tickets, answers the easy ones from the help-centre with citations, and routes everything else to the right human team with relevant customer context attached.
external
Funnel qualifies inbound demo requests in 90 seconds — enriches the lead, routes to the right AE based on territory and ICP rules, and sends the prospect a personalised calendar link before competitors get a sniff.
external
Endorse takes inbound policy endorsement requests, validates the change against policy rules, computes pro-rata premium adjustments, and generates the endorsement document in your firm's template.
external
Quill drafts the quarterly investor letter — pulls performance from the fund admin, writes the standard market commentary in your firm's voice, and leaves the strategic narrative for the GP to add.
external
Notify takes first-notice-of-loss reports across phone, email, and broker channels — runs the structured intake conversation by loss type, validates cover, and creates the claim file for the human adjuster.
external
Compass generates quarterly client portfolio reports — pulls positions, computes time-weighted return vs benchmark, and flags accounts needing advisor narrative. Renders branded PDFs and files them to client folders.
external
Ledger pre-screens new loan applications — checks document completeness against per-product checklists, chases applicants for missing items, and hands a clean file to the human credit officer.
external
Sentinel triages daily transaction-monitoring alerts, auto-closes obvious false positives with documented rationale, and hands the rest to investigators with a draft narrative. Works against any rule-based monitoring platform.
external
Reckon reconciles daily payment clearing files against the core banking ledger, flags exceptions over threshold, and drafts the regulator-mandated incident notes for human sign-off. Built for retail banks and SME lenders running NPP, BPAY, and card-scheme settlements.
external
Vital submits insurance pre-authorisation requests for elective procedures — pulls clinical evidence from the medical records system, matches each payer's required format, chases approvals daily, and routes rejections to the human authoriser with a draft appeal.
external
Echo reads every new public review across review platforms, drafts an on-brand response, and surfaces anything legally sensitive for human-only handling. Produces weekly themes summaries for the product team.
external
Pulse compiles your team's weekly reporting pack — pulls metrics from the relevant systems, chases owners for narrative inputs, lays it out in your standard template, and sends to the distribution list on schedule.
external
Bid drafts procurement RFPs from the requesting team's brief — pulls standard sections from your template library, scopes evaluation criteria to the category, and queues for human procurement review before issue.
external
Lingua handles multilingual content needs — translates customer messages into your team's working language, drafts replies in the customer's language, and localises marketing copy respecting regional conventions.
external
Oracle is your team's internal helpdesk — reads every employee question across chat and email, answers from your policies, handbooks, and SOPs with citations, and routes anything genuinely ambiguous to the right human.
external
Welcome runs new vendor setup end-to-end — sends the onboarding pack, validates returned documents (insurance, banking, tax forms), creates the vendor record in your AP system, and only escalates exceptions to procurement.
external
Verify validates KYC documents during onboarding — checks ID document authenticity, beneficial ownership disclosures, address proofs against your firm's policy checklist, and escalates anomalies to compliance with a clear written reason.
external
Tally processes employee expense submissions — reads receipts, matches to categories and policy limits, queues clean items for reimbursement, and routes anything outside policy to the human approver with a clear rationale.
external
Slate handles the entire scheduling dance — reads your availability and preferences, negotiates times across multiple participants, sends invites with the right agenda and links, and reschedules cleanly when life happens.
external
Scribe joins your video meetings via meeting bot, captures the live transcript, produces a structured summary with decisions and action items, and routes those action items to the right people.
external
Inbox keeps your work email under control — classifies every incoming message, drafts replies for the routine ones, and surfaces only the messages that genuinely need your attention. Works for any sector or role.
external
Concierge takes reservations across phone, email, and direct messages — confirms seating against the venue's reservation system, sends confirmations, manages the cancellation waitlist, and blocks staff family meals and private events.
external
Hearth takes inbound rental applications across portals and direct channels, validates ID/income/rental-history docs, runs standard background checks, scores against landlord criteria, and presents a ranked shortlist with rationale.
external
Press produces the weekly newsletter — gathers content from contributors against the deadline, lays out in your brand template, runs link/image/personalisation pre-send checks, and pulls post-send engagement metrics.
external
Cadence takes the approved content calendar, adapts each post for each platform's format and constraints, schedules at the optimal time per channel, and monitors for any platform-specific publishing failures.
external
Sift tags bulk discovery documents for relevance, privilege, and key issues against the case's tagging protocol — produces a reviewer-prioritised queue and maintains a complete tag audit trail for any later defensibility challenge.
external
Clause takes inbound contract review requests, classifies them by type and urgency, runs conflict-of-interest checks, and routes to the right practice group with prior matter context already attached.