Overview
Blindspot is a browser-based English learning tool built around one observation: learners repeat the same mistakes because they never internalise the rule behind the error. Rather than functioning as a conventional spellchecker or a generic grammar-correction engine, the platform positions itself as a diagnostic and practice layer. It detects errors in a learner's own writing, explains why each correction is required, and then converts recurring patterns into targeted drills and vocabulary cards. The product lives at blindspot.study and ships as a freemium web application paired with a Chrome extension.
The gap it targets is well documented in self-directed language learning. Grammar-correction tools have become commoditised, but most deliver a corrected sentence without context. The learner accepts the fix, moves on, and reproduces the same structural error a week later. Blindspot's differentiator is the explanatory loop: every correction is paired with the rule governing it — the sample on the homepage corrects "didn't saw" to "didn't see" and then states that "didn't" already marks the past tense, so the following verb stays in base form. Clusters of similar errors are subsequently grouped into a shared grammar topic that can be practised.
Positioned within a crowded edtech and AI-writing field, Blindspot does not attempt to compete with general-purpose writing assistants aimed at native speakers. It treats the learner's own output — emails, messages, work notes, diary entries — as the curriculum. Interface localisation into Russian alongside English suggests an initial go-to-market focus on Russian-speaking learners of English, a segment that competing tools frequently treat as an afterthought.
Key Features
AI Error Detection with Rule-Level Explanations The core of the product is a text checker that does more than flag a mistake. Each flagged string is presented with a correction and a short rationale, as demonstrated by the homepage example that annotates "didn't saw" and explains the base-form rule that applies after "didn't". This explanation layer is the principal differentiator against proofreading tools that stop at the correction itself.
Personalised Grammar Practice Modules Errors are not treated as isolated events. The platform bundles multiple instances of the same underlying rule — three sample sentences containing "didn't saw", "didn't went" and "didn't understood" are grouped under a single topic, "Past Simple: negative sentences", with a concise rule summary. Learners then answer generated multiple-choice questions such as "She didn't ___ to the meeting", drawn from the same pattern, which closes the loop between diagnosis and retention. A dedicated section for grammar practice hosts the wider exercise library.
Vocabulary Builder with Collocations, Synonyms and Flashcards Words surfaced during writing checks or browsing can be stored in a personal dictionary. Entries are considerably richer than a simple translation: the sample entry for "endeavour" includes noun and verb senses, definitions, an example sentence, a collocation list ("make every endeavour", "creative endeavour") and synonym pairs (attempt · effort; try · strive). Flashcard review follows afterwards, and nothing is added automatically — the learner confirms each suggested word, which keeps the dictionary focused.
Chrome Extension for In-Browser Capture and Checking A browser extension allows word saving and writing checks without leaving the current page. Selecting a word and clicking a plus icon pushes it into the vocabulary list with a translation, while a second mode checks a draft directly in the browser and returns corrected text alongside grammar notes. This reduces the friction that normally separates reading from deliberate study, and further detail is available on the Chrome extension page.
Free English Level Test A standalone level test is offered as a placement tool, addressing the "not sure where to start" problem before a learner commits to a practice track. It also establishes a baseline against which later improvement can be measured, which is a useful feature for learners who lack external feedback.
Bilingual English/Russian Experience The interface ships in both English and Russian, with hreflang alternates declared for each locale. The writing examples map Russian phrasing ("готовился") to English equivalents ("was preparing / was studying"), indicating that the vocabulary system is designed to operate from a learner's native language rather than English-to-English only. A vocabulary section collects saved items and review material.
Editorial Blog and Reference Pages Dedicated sections for grammar, vocabulary and writing sit alongside a blog, giving the product a reference dimension beyond its interactive tools and providing a route for organic discovery.
How It Works
The typical user journey begins on the homepage, where a textarea accepts between 30 and 300 characters — an email, a message, a work note or a few sentences about the day. Pressing the check button triggers authentication: if the visitor is not signed in, a window opens to continue with Google. The draft is retained in the current browser tab rather than being discarded.
After signing in and selecting a native language where necessary, the text is submitted to OpenAI for analysis. The result opens in a feedback view that separates three concerns: the corrected sentence, the grammatical reason behind the change, and a suggested vocabulary item extracted from the text. The learner chooses whether to save that word, so the dictionary grows deliberately rather than automatically.
Accumulated mistakes feed a second layer. When several errors share a root cause, they are aggregated into a named grammar topic with a short explanation and a set of generated questions. Answering those questions produces new performance data, which in turn refines which topics surface next. Once accounts exist, the checker within the app accepts texts up to 5,000 characters, roughly sixteen times the landing-page limit.
For continuous use, the Chrome extension intercepts the workflow at the point of reading. A word highlighted in an article can be captured with a single click; a paragraph being drafted can be checked without switching tabs. The overall architecture is therefore a closed cycle of write, diagnose, explain, drill and capture.
Use Cases
Daily writing journals. A learner who writes a short English entry each evening receives immediate corrections and, more importantly, a running tally of which rules keep failing. Over a month, the grammar practice queue becomes a personalised syllabus rather than a random exercise set.
Business and workplace correspondence. Professionals preparing emails, Slack messages or client notes in a second language can paste a draft before sending it. The explanations help distinguish between errors that merely look untidy and those that change meaning, which matters more in a workplace context than in casual writing.
Vocabulary expansion for intermediate learners. A learner who already communicates adequately but lacks range can use the word-saving flow to build a dictionary of collocations and synonyms rather than isolated translations. Because entries include usage examples and collocation lists, the saved material is directly reusable in production.
Native-language-to-English bridging for Russian speakers. The bilingual interface and native-language word mapping let a Russian-speaking learner start from a phrase they already know and receive English equivalents with contextual notes, which is faster than working exclusively in a monolingual dictionary.
Placement and self-assessment. Learners who are unsure of their level can begin with the level test, establish a baseline, and then use the checker and practice modules to track progress against concrete error categories rather than vague impressions.
Pricing & Value
Blindspot operates on a freemium model. Creating an account provides free access to text checks and personalised practice with no card required, and the homepage states explicitly that payment can be chosen later for additional practice. No specific price points are published on the landing page, which makes a precise value assessment difficult without entering the product.
The practical value proposition rests on the explanation layer. Free grammar checkers already exist in abundance, so the chargeable element is the diagnostic and practice loop built on top of corrections. Whether that justifies a subscription depends on how heavily a learner uses the practice modules. For occasional proofreading the free tier is likely sufficient; for consistent daily study the paid expansion is the more plausible purchase. The absence of transparent pricing remains the most notable friction point in an otherwise clear funnel.
Final Verdict
Blindspot addresses a genuine weakness in the AI writing-tool market: correction without comprehension. By pairing every fix with a rule and then reusing those rules as practice material, it converts a passive checking habit into an active learning loop, and the Chrome extension extends that loop into ordinary browsing. The bilingual English/Russian design and the depth of the vocabulary entries — collocations, synonyms and flashcards rather than bare translations — give it a more considered feel than most lightweight checkers.
Areas for improvement are equally clear. Detailed pricing is not disclosed on the public site, the landing-page checker is capped at 300 characters, and the reliance on OpenAI means submitted texts leave the browser and are stored server-side, which privacy-conscious users should weigh against the convenience. The product is nonetheless a coherent, well-scoped tool for intermediate learners who want their own writing to drive their study rather than generic exercise books.
