Roadmap — last updated 10 July 2026

What's coming. And what we're still figuring out.

Everything that's already live is on the features page. This is the road ahead — and it's the contract: if a feature is here with a quarter target, that's when it ships. If it slips, the next update will say so — we won't quietly move it.

Now building — next 1–2 quarters

In progress. Design partners get these as we ship.

CommittedQ3 2026

Live bank feeds (Account Aggregator)

Automatic daily transaction sync from your bank — no more CSV imports. RBI-approved AA framework: the Setu adapter, FI-data mapping, and reconciliation engine are all built and tested — code-complete.

Blocker — Only the Setu enterprise contract (bank-side, not code) remains before feeds go live.
CommittedQ3 2026

Multi-user real-time collaboration

Two users in the same voucher / invoice see each other live — presence + cursors + last-write-wins field sync, now on both the invoice form and the voucher receipt/payment form (edits broadcast live; voice prefills sync too).

Blocker — Production-load testing of Supabase Realtime at scale remains before GA.
CommittedQ3–Q4 2026

Self-serve plugin publishing

The app/integration marketplace (publish → review → install, with pricing) is already live. Declarative plugins (validators, config) need no sandbox and are self-serve-safe now. For arbitrary-code plugins the sandbox is designed (docs/PLUGIN-SANDBOX-DESIGN.md) and its capability bridge — deny-by-default permission gate + call/output budgets + clone-safety — is built + unit-tested; the isolate runtime (QuickJS-wasm) + escape-test suite are the remaining phase. Code plugins are invite-only review until then.

Blocker — For arbitrary-code plugins: the sandbox isolate runtime + escape-test suite + a security review. Declarative plugins are self-serve-safe today.
CommittedQ3 2026

GSP / ESP go-live

Direct in-app e-invoice IRN, e-way bill, and DSC e-signing. Code paths are built behind flags.

Blocker — GSP + ESP commercial contracts.

Planned

Committed direction, dates subject to design-partner feedback.

PlannedQ4 2026

Group consolidation + multi-currency depth

For Indian businesses with overseas subsidiaries. The /dashboard/consolidation screen now renders the consolidated Schedule III balance sheet + basis-of-consolidation notes and exports a PDF; the pipeline is unit-tested end-to-end (95 tests) and balances including investment + NCI cases (pre-acquisition equity elimination posted). The live path derives NCI for single-subsidiary runs; multi-subsidiary NCI + goodwill are now built + unit-tested end-to-end in the engine (a two-sub group with goodwill balances: assets = owners’ equity + combined NCI). The per-entity-TB → core-input mapper is built + tested too, so the whole chain (fetched TB → consolidate → assemble → statement) is proven pure. Remaining: the thin live-route DB fetch (it uses the v1 combine-engine today) + acquisition-cost capture in the UI for goodwill, and splitting NCI on current-year profit.

Blocker — All consolidation logic is built + tested; what remains is the live-route per-entity fetch + acquisition-cost UI fields, verified in the running app.
PlannedQ4 2026

Plugin API v2 (blocking validators)

Live: a safe declarative rule DSL + a validate:save permission + a host aggregator, now enforced on the invoice and voucher save routes (receipt / payment / journal) — a registered validate:save plugin blocks a non-compliant entry with its message, and it’s a no-op when none is registered. Prototype-safe, no regex/ReDoS, depth-capped, fail-open; unit-tested. Validators are interpreted data, not executed code — no sandbox needed. A WASM sandbox is still required only for plugins that run arbitrary code.

Blocker — Only the WASM sandbox isolate (for arbitrary-code plugins) + its security review remain — the validator surface is already live.
PlannedQ1 2027

24×7 support tier

Currently business hours (Mon–Sat, IST). Adding overnight WhatsApp + email coverage during the March + September filing peaks.

Blocker — A support hire — recruiting + onboarding overnight staff. Not a code task.
PlannedQ1 2027

Customer reference programme

Public case studies from the first design-partner CAs — with permission and editorial review.

Blocker — Needs the first design-partner CAs live + their written consent to go on record. Not a code task.

Evaluating

Ideas we believe in but have not committed dates for. Tell us if you need any of these.

EvaluatingLater

Tally drop-in replacement

A migration + workflow that replaces Tally entirely for a given firm size. Today we coexist (two-way sync); this would replace.

Blocker — A committed multi-month build — deliberately not scheduled yet (it’s in Evaluating).
EvaluatingLater

Voice-driven bookkeeping

Mobile voice → journal entry is live (Hindi / English + 6 more Indian languages, Whisper / Sarvam ASR). NER party resolution is built + tested, and in-browser voice capture is now wired on the web voucher form too — speak a receipt/payment and it prefills amount, mode and narration and auto-matches the party. Still evaluating: validating the extra languages with native speakers.

Blocker — Native-speaker validation of the additional Indian languages — the core flow is already live.
EvaluatingLater

On-prem / private-cloud option

For larger CAs and groups uncomfortable with cloud. Same code, customer's infra. Compliance-driven.

Blocker — A specific enterprise commitment + compliance (SOC2-type) packaging — demand-driven.

Need something that's not on here?

We're running a design-partner programme right now. The first 10 CAs / firms get to set the order in which the Q3–Q4 items ship. If you've got something specific you need, tell us — we'll either move it forward, add it to evaluating, or honestly say no.

Talk to foundersSystem status →