Blits.ai Platform Update: Five Big Additions from the Last Six Months

We ship continuously, so it's easy to miss how much an agent can do now that it couldn't at the start of the year. This is a plain-language roundup of the five biggest additions from the last six months, what each one is, and what it changes for your customers. No engineering detail, just the parts you'd notice.

At a glance:

1. Voice emotions: agents express emotion in both the voice and the digital humans.
2. Rich interactive cards: a growing library of in-chat components, not walls of text.
3. Agentic tasks: the agent can schedule and defer actions, and act on triggers.
4. Agentic Pay: take payments inside the conversation, built to PCI-DSS Level 1.
5. Response feedback & self-learning: capture ratings, and improve from them.

1. Voice emotions in avatars and speech

Agents can now express emotion, warmth, calm, empathy, a firmer tone when it's warranted, and a single cue drives both the digital human's expression and the spoken voice. A reply can shift tone partway through, opening calm and landing firm. It's a setting on our agents, not something you hand-write into a prompt, and it works across the voice engines we support, including Gemini, ElevenLabs, Azure, and Cartesia.

We also fixed a rough edge in the same release: emotion cues could occasionally get read aloud or show up as stray characters. That's now handled cleanly on every channel.

Why it matters: a voice that sounds like it means what it says builds trust, especially in the moments that count, a complaint, a declined payment, a sensitive request. For more on why voice quality is hard to get right, see dialects and the voice-quality gap.

Availability: live.

2. Rich interactive cards Library 

An agent now renders interactive components right in the conversation: swipeable flight and hotel carousels, product and loyalty cards, receipts, a spending graph, a secure field for a one-time code, and native WhatsApp buttons and lists instead of a menu pasted into a message. Customers scan and tap instead of reading and typing.

These cards are also structurally reliable: the agent builds each one as a validated action rather than free-form text, so whole classes of "the card didn't render" can't happen.

Why it matters:

• Customers act instead of read — better conversion, clearer flows.
• The right card at the right moment (a payment, a receipt, a choice) inside a normal conversation.
• Works across mobile, digital human, and WhatsApp, using each channel's native elements.

Availability: on by default for new agents.

3. Agentic tasks

Agentic tasks let an agent take on work that happens later, on its own — a follow-up if something hasn't arrived, a check that runs every morning, or an action kicked off the moment a specific email lands or a file is uploaded. Before it commits, the agent confirms what it's going to do and keeps a summary of what's queued, so nothing slips.

Why it matters: the agent works proactively instead of only reacting, while staying bounded and reviewable.

Availability: Live with first customer on travel use-case

4. Agentic Pay

Agents can now take a payment inside the conversation — send a payment link, confirm a transaction, and show the result as a clean card — without your business ever touching raw card data. Card details are tokenized straight through the payment provider, the platform never stores a card number or a security code, and the sensitive part runs in its own tightly-scoped environment. We built it to the PCI-DSS Level 1 standard and wired in multiple payment providers, including regional ones like Qi in Iraq, so it works in the markets our customers operate in.

Why it matters: these are compliance-grade rails that let money move inside a conversation, the kind a risk team can actually approve. For the practical side, see how to let an agent transact without losing control.

Availability: Live

5. Response feedback and self-learning

Customers can now give any reply a thumbs up or down and add a comment. That feedback is captured, tied to the exact response it's about, and collected in one place your team can review, so you can see where the agent lands well and where it's missing. The signal also feeds back into the agent, so it improves from real conversations instead of waiting for someone to rewrite a prompt every quarter.

Why it matters: you get a clear, structured read on quality — and an agent that gets better on its own, which is the self-learning shift we called the next frontier in Johannesburg.

Availability: Live

The short version

That's the last six months: an agent that can convey emotion, show things instead of describing them, act on its own schedule, take a payment your compliance team signs off on, and learn from whether it got it right.

None of it shipped as a big-bang launch, each piece went live when it was solid enough to trust with a real customer on a real account. If any of it maps to something you're building, we'd like to hear about it.