Financial wellbeing

On Blits.ai the coach is an AI agent whose numbers come from custom functions that call the bank's data and forecasting services, or from a SQL knowledge base the agent can query, so the model explains figures rather than calculating them. Guidance content sits in a knowledge base with hybrid retrieval. Actions such as creating a savings goal run as flows or agentic workflows with a tool execution policy and human in the loop confirmation above a threshold.

The coach lives in the mobile app through the REST or WebSocket API channel, with rich cards for balances, charts and insight panels and optional voice. Guardrails enforce the advice boundary and block sales language in coaching answers, PII masking protects transaction data, test suites check the boundary cases on every change, and analytics track use and satisfaction. The platform is model agnostic. Run Blits.ai in our cloud in the EU, Qatar, Singapore, the UAE, the US or Australia, in your private cloud, or fully on premises.

On Blits.ai this is an AI agent on voice (telephony with real time streaming speech recognition and synthesis, and call transfer), SMS, WhatsApp and email, with the regulated steps (identity check, disclosures, arrangement offer) built as deterministic flows and DTMF input where customers prefer the keypad. Custom functions read arrears data and write arrangements to the servicing system, and payments go through payment links from the payments service, so card numbers never enter the conversation; on voice, DTMF capture with the sensitive data flag keeps digits out of the transcript, and in text channels credit card number detection and tokenization and PII masking at the gateway tokenize or redact card numbers a customer types.

Sentiment analysis and guardrails help detect distress, and human handover sends hardship and vulnerability cases to a specialist with a conversation summary, including live takeover of a voice call. Agentic tasks schedule follow ups such as a check after a promised payment date. Test suites replay hardship scenarios on every change, and analytics show interactions, satisfaction and sentiment. The platform is model agnostic and supports several languages, including Arabic.

On Blits.ai each trigger is an agentic workflow started by the bank's systems through an API token or on a schedule, with the customer, the reason and any pre approved offer passed in as data. Blits.ai can send the first message itself on the email channel, which works inbound and outbound. For outreach in the bank's own app, the app starts the contact and hands the conversation to the AI agent through the REST or WebSocket API channel.

For WhatsApp, SMS and voice, the bank's own messaging or dialler platform sends the first message or places the call, and the agent answers the replies on the WhatsApp, SMS and voice telephony channels. In every case the agent answers from a knowledge base of approved product content and acts only through custom functions that call the bank's APIs, behind an authentication block.

A flow fixes the opening (AI disclosure, reason for contact, opt out) and the consent checks, while the GDPR toolkit handles consent gating and data removal. Human handover routes complaints and hardship to the contact center, guardrails stop the agent from inventing offers or asking for secrets, and test suites replay each trigger's conversations before launch. The workflow run history with analytics shows outcomes per trigger for comparison with the control group.

On Blits.ai the conversation runs in an AI agent that gathers goals and facts, with a flow for the fixed parts of the fact find and custom functions that call the firm's planning engine through its API, so every projection comes from the engine. A knowledge base holds the approved explanations of assumptions and concepts, and structured output keeps numbers in separate fields that a custom function can compare with the engine response.

The advisor version runs in Microsoft Teams or inside internal tools through the REST API channel; a client version can use the web chat channel, or the API channel inside the firm's own app, with rich cards and charts to compare scenarios. Guardrails stop product recommendations and hand complex cases to a human through human handover, and test suites replay known scenarios on every change. Multi language support covers markets where clients plan in their own language.

On Blits.ai the intervention is an AI agent called from the payment flow through the REST API or WebSocket API channel, or shown in the embedded chat widget, with the payment context and the risk band passed in. A flow fixes the regulated steps per band (warning, questions, hold offer) and the agent handles the open conversation. Custom functions call the bank's scoring service, the payee check and the payment hold API; a knowledge base holds the approved scam education content per typology.

Held payments go through an agentic workflow with human in the loop approval, so a scam specialist approves or rejects release with the conversation attached, and every step lands in the run's audit trail. Human handover passes the customer and the conversation history to a specialist in chat or on a voice call, so they do not repeat the story.

Guardrails and PII masking keep account data out of prompts, test suites replay scam scripts (including coached answers) on every change, and analytics track cancellations and handovers per typology. The platform is model agnostic. Run Blits.ai in our cloud in the EU, Qatar, Singapore, the UAE, the US or Australia, in your private cloud, or fully on premises to meet data residency requirements.

On Blits.ai the decisioning itself stays in the bank's next best action engine; the platform delivers and explains its decisions. An AI agent in the app or on WhatsApp calls the engine and the loyalty platform through custom functions, shows offers and points with rich cards (loyalty points, credit cards, product recommendations) and completes redemptions through flows with confirmation.

Offer terms and rewards rules sit in a knowledge base so explanations are grounded. Reward values come from the loyalty platform, and guardrails help prevent the agent from inventing values or pushing credit to customers who mention financial difficulty, while human handover routes those conversations to people. Analytics track interactions, satisfaction and sentiment, and test suites check the suppression rules before each release. The platform is model agnostic.