The full framework is published as a research paper at taralabs.ai/paper/basic-paper. To our knowledge, it is the first research paper in India to treat credit score improvement as its own machine learning problem. The paper introduces a new field of study — Behavioural Credit Analytics — focused on turning AI insight into measurable consumer benefit.
The 300 Million Indians the System Forgot
India has over 300 million credit-active consumers. Most of them sit in the middle. Not chronic defaulters. Not prime borrowers. They sit in a near-prime band where a 50-point swing in their credit score is the difference between a loan at 12% and a loan at 22%. The difference between approval and rejection. The difference between buying a two-wheeler this month or waiting another year.
These are gig workers in Bengaluru. Self-employed shopkeepers in Indore. Salaried professionals in Tier 2 and Tier 3 cities whose financial lives don’t fit the templates that credit scoring models were built for. They check their score every month on an app. They see the number. They feel the weight of it. And they have nowhere to turn for a real plan.
“Credit scores answer one question well — where are you? They were never designed to answer the next one: how do you move? We see this every day. People come to us with scores of 580, 620, 650 — not because they’re irresponsible, but because nobody has shown them how to move. Until BASIC, nobody was giving these borrowers a real, personalised plan to fix their financial future. That’s what we set out to change,” said Anupam Acharya, Co-founder and CEO of GoCredit.
What’s Different About BASIC
Until now, an Indian consumer with a low credit score had three options. Generic advice from public websites. Paid credit counsellors who couldn’t see their actual report. Rule-based simulators that gave the same answer to everyone, regardless of profile.
None of these produced a real, personal plan. None of them could tell a specific borrower with a specific credit profile what to do, in what order, to gain the most points in the least time. That is what BASIC does — and as far as we know, no one has done it before in India.
So what makes BASIC work? Most credit models look only at where you are today. BASIC looks at where you are, where you’ve been, and how fast you’re moving — for every part of your credit profile. Trained on the credit histories of more than a million Indian borrowers over four years, BASIC can predict your next-month credit score with an average error of just 16 points. On the 300–900 credit score scale, that is under 2.2% off. Standard deep-learning models trained on the same data are about four times less accurate.
“A credit score isn't just where you are today - it's where you're heading. Most score simulators miss this: they look at your current credit attributes in isolation, when bureau algorithms are actually responding to how those attributes are changing over time. We trained BASIC on the rate of change alongside the current value, which is the temporal structure the bureaus actually reward. That's why the recommendations move scores instead of just describing them,” said Pranav Murali, Co-founder and Lead AI Engineer at GoCredit.
Already Used by Over 2 Lakh Indians
This is not a research preview. BASIC is already live on the GoCredit Android app, where it powers a feature called Credit Score Boost — used by over 2 lakh Indians so far.
Among the 50,000 monthly active paying subscribers, 10,000 — one in five — have improved their credit scores by 50 points or more within two months of actively following BASIC’s recommendations.
For these 10,000 borrowers, that 50-point gain is not abstract. For someone near the prime threshold, it can be the difference between a loan rejection and an approval. It can mean an EMI at 12% instead of 22%. For many, it is the first time their credit score has moved in their favour in years — and the first time anyone has shown them how to make it happen.
“We meet borrowers every day who tell us they’ve been trying to fix their score for two or three years and getting nowhere. Two months later they show us their app and the number has moved up by 60, 70, 80 points. That moment — when a borrower realises they’re not stuck, that the system isn’t a black box anymore — that’s why we built this,” said Anupam Acharya, Co-founder and CEO of GoCredit.
Now Opening to Lenders: The ScoreBoost API
TARA AI Labs is also announcing today that the same engine is now available to financial institutions through the ScoreBoost API. The API is live with select pilot partners.
For the first time, a lender selling a loan, credit card, or BNPL product can show a borrower — right at the point of sale — how that product will affect their credit score over the next 6 to 12 months. A consumer comparing two credit cards can see, before signing up, which one is more likely to help their score and which one might hurt it. A lender offering a secured loan can show the borrower exactly how the product fits into their improvement journey.
“Indian lenders today sell loans without showing customers the consequence on their credit profile. That information asymmetry has been one of the most under-discussed problems in retail lending. ScoreBoost API closes it. A lender can integrate our engine and tell a borrower: here is what this product will do to your credit score. It changes the conversation from ‘do you qualify?’ to ‘is this the right product for where you’re trying to go?,” said Shubham Arawkar, Co-founder and CTO of GoCredit.
ScoreBoost API is currently in pilot with select NBFCs and digital lending platforms. Broader availability will be announced later in 2026.
Privacy and Regulatory Architecture
BASIC works only on consumer-consented credit data, accessed through official APIs under the consumer-initiated credit information report framework of the Credit Information Companies (Regulation) Act, 2005. All data handling also complies with India’s Digital Personal Data Protection Act, 2023. No bank account access, no SMS reading, no invasive data collection — by design.
BASIC is deployed strictly as a consumer-facing advisory system. Its outputs are not routed — through any production pathway — into GoCredit’s lending decisions, underwriting models, or partner lenders’ scoring engines. The advisory model and the lending stack are operationally separated systems.
“A model trained to help consumers improve their scores shouldn’t simultaneously be used to judge them. We enforce that separation operationally — BASIC’s outputs have no production pathway into any lending decision engine, ours or anyone else’s,” said Shubham Arawkar, Co-founder and CTO of GoCredit.
Availability
-
Credit Score Boost is available now on the GoCredit Android app at INR 199/month.
-
ScoreBoost API is in pilot with select financial institutions; institutional partners can request access at research@taralabs.ai.
-
The full BASIC research paper is available at taralabs.ai/paper/basic-paper.
About TARA AI Labs
TARA AI Labs is the deep-tech research division of GoCredit, focused on responsible AI for credit and financial empowerment in emerging markets. The lab publishes research on credit modelling, applied AI on financial data, and consumer-centric AI design. taralabs.ai.
About GoCredit
GoCredit is an AI-powered personal finance platform helping Indians access credit and improve their financial health, with over INR 120 crore in loan disbursals facilitated across 50+ RBI-regulated lending partners. gocredit.money.
