The objection you answer well is the case the Tribunal never sees.
The Kenyan Tax Procedures Act gives the Commissioner 60 days to respond to a notice of objection. Inside that window, an objection response that engages the taxpayer's specific evidence resolves a meaningful share of cases without escalation. The authority that handles objections well loses fewer cases at TAT — and loses less staff time to litigation.
The Tax Appeals Tribunal docket is the visible part of the dispute resolution function. The invisible part — the part that determines how much of the docket exists in the first place — is the objection process inside the authority. A taxpayer who receives an additional assessment has 30 days to lodge a written objection with the Commissioner. The Commissioner has 60 days to respond. Inside that 90-day window, two things can happen. The Commissioner can vacate or amend the assessment in light of the objection. Or the objection conversation can resolve the matter without escalation, sometimes through Alternative Dispute Resolution with its further 90-day timeline. Only the cases that survive this funnel reach the Tribunal.
Most revenue authorities are operating the funnel as a paper exercise. The objection arrives. An officer is assigned. The officer reviews the file, considers the taxpayer's arguments, drafts a response. The response either confirms the original assessment with reasons, amends it, or vacates it. The work product is a letter. The quality of the letter — its engagement with the taxpayer's specific evidence, its articulation of why the Commissioner's position holds, its identification of points where the Commissioner concedes — determines whether the case escalates to TAT or closes.
The pressure on this function has grown faster than the function. KRA's expanded use of validation engines means more assessments are being raised. The 2024-2026 TAT decisions show a docket that is larger and more technically complex than five years ago. FIRS's Tax Reform Acts 2025 have produced a similar volume increase in Nigeria. RRA's EBM-driven VAT assessments generate a steady objection flow. SARS's pre-population programme produces queries on auto-assessments that the legacy SARS Disputes Unit was not sized for. The objection function is the bottleneck — and the place where authorities lose ground that they took at the assessment stage.
Every case that reaches the Tribunal is a case the objection process failed to close. The economics of the authority's litigation budget run through that fact.
The work an objection response must do is precise. It must engage with the taxpayer's specific objection grounds — not the generic categories of objection, but the actual arguments and evidence the taxpayer's representative has put in front of the Commissioner. It must explain why the Commissioner's original reasoning holds against those arguments, or why it gives ground on some points and not others. It must produce a defensible record for the file in case the matter escalates. And it must do all this inside 60 days, across a portfolio that may include several dozen active objections per officer at any one time.
The natural response is to standardise. Standardised objection responses save officer time and produce consistent quality. They also lose at TAT. The Tribunal is not impressed by a response that quotes Section X of the Income Tax Act without engaging with what the taxpayer actually said. The standardisation that protects officer hours is the same standardisation that produces letters the Tribunal reads and dismisses.
Where each sits.
Akki sits between the authority's case management system and the AI tools the objection function uses. When an objection is lodged, the case file already contains the assessment, the reasoning trail behind it, the third-party data the assessment was built on, and the taxpayer's filings. The taxpayer's objection — the notice itself, the statement of facts, the supporting documents — is added to the file. The officer is working from a structured record rather than reconstructing the picture from emails.
Solva is the layer that produces the response. When the officer reviews the objection, Solva structures the reasoning the response must do. It restates the question — does the taxpayer's objection, taken at its best, change the Commissioner's position. It surfaces what is in the taxpayer's evidence, what the original assessment reasoning rested on, and what would have to be true for the Commissioner to amend or vacate. It triangulates between the taxpayer's argument, the data the authority holds, and the authority's prior treatment of similar arguments at TAT. It produces a draft response in which the engagement with the taxpayer's specific points is explicit. Where the objection has merit on one point and not on others, the draft says so, with reasons. Solva's refusal behaviour does specific work for the objection function. When the Commissioner's original assessment was thinly reasoned — when the file shows the assessment was raised on a weight of third-party data without proper triangulation — Solva refuses to draft a confirmation. It produces, instead, a recommendation to amend or vacate. This is the moment the authority's economics turn. The officer is no longer escalating to TAT a case that the Commissioner will lose. The objection closes inside the 90-day window. The Tribunal's docket shrinks. The cases that proceed are the ones the authority should fight.
SyniSense governs the data layer for the objection process. The taxpayer's objection often introduces new evidence the authority did not hold at the assessment stage. That evidence is sensitive and is held under data protection obligations. SyniSense ensures that when external AI is used to review the new evidence — comparing the taxpayer's supplied bank statements to the authority's third-party data, validating cancelled-invoice claims against eTIMS — the identifiable fields are anonymised at the boundary. The audit receipt records what was reasoned about and on what data.
The objection officer's workday changes. The officer is reviewing reasoning drafts the AI has done, not drafting from scratch. The cases that hold under the Commissioner's own scrutiny get strong responses. The cases that do not hold get conceded inside the 60-day window.
The Alternative Dispute Resolution conversation changes. ADR works when both sides understand the file. When the objection response is well-reasoned, the taxpayer's representative arrives at ADR knowing what the Commissioner thinks and why. Settlements are reached on substance rather than on the cost of litigating.
The Tribunal docket changes. Cases the authority would have lost are not appearing in front of the Tribunal because the Commissioner already vacated them. Cases the authority will win are appearing with files that show the reasoning end-to-end. The win rate rises. The litigation budget reallocates from defending weak files to pursuing the cases that matter.
And the taxpayer relationship changes. The taxpayer who lodges a competent objection and receives a competent response engages with the system differently than the taxpayer who lodges an objection and receives a template letter.