{"name":"Fidence-AI","url":"https://fidence-ai.com","description":"For whoever runs the model layer. Fidence-AI trains, evaluates, scores and delivers a model for every team that asks, in minutes, on your own data, for your people and agents.","updated":"2026-11-09","howToRead":["Figures for the old way come from published research and list prices, with arithmetic shown.","Worked examples are arithmetic on starting numbers a reader can change. They are not customer results."],"audience":"CIOs, CDOs and heads of data who serve models to teams inside the company.","modelLayer":[{"job":"Train","today":"3.3 to 4.5 months on average for data science, MLOps and low-code platforms to go live (G2, 2026), and each request still takes your team's weeks.","whatTheFactoryDoes":"Connects to your data, profiles it and trains the model. The same process for a yes or no, a category, a number or a forecast.","result":"Minutes","resultNote":"From connected data to a tested model"},{"job":"Evaluate","today":"Your board wants proof of return, and most teams lack a mature way to measure it.","whatTheFactoryDoes":"Tests every model on records held out of training, grades it Decision-grade, Directional or Not usable, in words the team that asked can read.","result":"Graded","resultNote":"Decision-grade, Directional or Not usable"},{"job":"Score","today":"A scoring pipeline to build and keep alive for every model your teams use.","whatTheFactoryDoes":"Applies the saved model exactly as saved, with no language model in the path. It checks itself against rows it has scored before, and writes nothing if they do not match.","result":"Under 1¢","resultNote":"Of computing per million rows"},{"job":"Serve","today":"Endpoints to keep running or batch jobs to schedule, for every team that asks.","whatTheFactoryDoes":"Keeps saved models ready and runs them as batch jobs when you ask. Scored rows land wherever you define, for your people and agents to read.","result":"One request","resultNote":"For the next batch, with no pipeline to maintain"}],"summary":[{"promise":"Fast","measure":"From data to a tested model","before":"Weeks per request","after":"Minutes","note":"Against the weeks a request takes your team today."},{"promise":"Cheap","measure":"Scoring a million rows","before":"A fee per row","after":"Under 1¢","note":"The saved model's computing, against a language-model call or an always-on endpoint."},{"promise":"Always ready","measure":"The next batch","before":"A pipeline to maintain","after":"One request","note":"Saved models stay ready and run as batch jobs when you ask, with no pipeline to maintain."},{"promise":"Yours","measure":"What you are left with","before":"A fee per answer","after":"Your model","note":"Trained only on your data and kept in storage of its own, not weights on a provider's platform."}],"outcomes":[{"id":"churn","name":"Churn","internalClient":"Customer success","question":"Who is about to leave?","pain":{"figure":"88% to 84%","text":"is how far gross revenue retention fell, in a benchmark of 342 software companies. Its authors call it “a market-level structural dynamic, not company-level execution.”","source":"Benchmarkit and Aleph, 2026"},"fix":"Trains a classifier on your own history of who left, and gives every current customer a probability and a high, medium or low band. The week starts with the list already sorted by risk.","addsToEveryRow":"A probability and a band per customer, tested on records held out of training.","workedExample":{"title":"The same number of calls, more of the customers who are leaving","saving":"$144,000","savingIs":"of revenue kept, a year","working":"8,000 of 50,000 customers leave each year. Contacting 5,000 in no particular order reaches 800 of them. Contacting the 5,000 the model ranks highest reaches 2,000. That is 1,200 more. If 10% of them stay because of it, that is 120 customers kept, worth $144,000 a year.","inputs":[{"id":"total","label":"Customers","value":50000,"min":100,"max":50000000,"unit":null},{"id":"rate","label":"Who leave each year","value":16,"min":1,"max":90,"unit":"%"},{"id":"share","label":"Customers your team can contact in a year","value":10,"min":1,"max":100,"unit":"%"},{"id":"top","label":"Leavers in the model's top tenth","value":25,"min":10,"max":100,"unit":"%"},{"id":"save","label":"Contacted leavers who stay because of it","value":10,"min":0,"max":100,"unit":"%"},{"id":"value","label":"Revenue per customer, a year","value":1200,"min":1,"max":10000000,"unit":"$"}],"assumption":"The top-tenth share and the save rate are starting figures. Replace them with what your own hold-out test shows.","isCustomerResult":false}},{"id":"lead","name":"Lead conversion","internalClient":"Sales","question":"Which leads will convert?","pain":{"figure":"43.6%","text":"was quota attainment across 252 software companies in a recent quarter, while sellers’ rating of inbound lead flow fell to its lowest recent reading.","source":"RepVue, 2026"},"fix":"Trains on the leads you already won and lost, and scores every open lead with its probability of converting. Reps work the pipeline from the top.","addsToEveryRow":"A probability per lead, tested on held-out records.","workedExample":{"title":"The same number of calls, more of the buyers","saving":"$420,000","savingIs":"more deal value in front of your reps, a year","working":"25 of 1,000 leads a month would buy. Working 250 in no particular order puts reps in front of 6 of them. Working the 250 the model ranks highest puts them in front of 13. That is 7 more buyers a month, or $420,000 a year in deals. It assumes a lead converts only if someone works it.","inputs":[{"id":"total","label":"New leads a month","value":1000,"min":10,"max":10000000,"unit":null},{"id":"rate","label":"Who would buy if worked","value":2.5,"min":0.5,"max":90,"unit":"%"},{"id":"share","label":"Leads your reps can work","value":25,"min":1,"max":100,"unit":"%"},{"id":"top","label":"Buyers in the model's top fifth","value":50,"min":20,"max":100,"unit":"%"},{"id":"value","label":"Value of one deal","value":5000,"min":1,"max":100000000,"unit":"$"}],"assumption":"The top-fifth share is a starting figure. Replace it with what your own hold-out test shows.","isCustomerResult":false}},{"id":"fraud","name":"Fraud","internalClient":"Risk","question":"Which transactions need a second look?","pain":{"figure":"$5.13","text":"is roughly the total cost of each $1 of direct fraud loss at US retail and ecommerce companies, in a survey of 513 risk and fraud leaders across the US and Canada.","source":"LexisNexis Risk Solutions, 2026"},"fix":"Trains on the cases you have already confirmed, and returns every new transaction with a probability and a risk band, flagging the riskiest. Reviewers open those first.","addsToEveryRow":"A probability and a risk band, tested on held-out records.","workedExample":{"title":"Less fraud through, and what each dollar of it really costs","saving":"$256,500","savingIs":"of fraud cost avoided, a year","working":"$500,000 of direct losses × $5.13 is $2,565,000 a year in total cost. If the model stops 10% of that fraud, and the other costs fall in step with it, that is $256,500 a year.","inputs":[{"id":"loss","label":"Direct fraud losses a year","value":500000,"min":1000,"max":10000000000,"unit":"$"},{"id":"multiple","label":"Total cost of each $1 lost","value":5.13,"min":1,"max":20,"unit":"$"},{"id":"stopped","label":"Of that fraud, what the model stops","value":10,"min":0,"max":100,"unit":"%"}],"assumption":"The multiple comes from published research. The 10% is a starting figure: replace it with what your own hold-out test shows.","isCustomerResult":false}},{"id":"ltv","name":"Lifetime value","internalClient":"Marketing and finance","question":"What is each customer worth?","pain":{"figure":"16 months","text":"is how long the median software company takes to earn back what it paid to win a customer. The bottom quarter takes 24 months or more; the top quarter, 6 or fewer.","source":"Aleph and Benchmarkit, 2026"},"fix":"Trains a regression model on what past customers went on to spend, and gives every customer an expected value. Budget and attention can follow the number.","addsToEveryRow":"An expected value per customer, tested on held-out records and checked again after it is saved.","workedExample":{"title":"A value for every customer, kept current","saving":"160 hours","savingIs":"of analyst time returned, a year","working":"4 refreshes × 5 days × 8 hours = 160 hours by hand. At ten minutes a run, 4 refreshes take about 40 minutes.","inputs":[{"id":"refreshes","label":"Times a year you refresh the values","value":4,"min":1,"max":52,"unit":null},{"id":"days","label":"Analyst days per refresh, by hand","value":5,"min":1,"max":60,"unit":"days"}],"assumption":"The days are your own estimate.","isCustomerResult":false}},{"id":"next","name":"Next purchase","internalClient":"Marketing","question":"What will each customer buy next?","pain":{"figure":"56%","text":"of shoppers often or very often receive irrelevant messages, and 80% are more likely to ignore brands that send them, in a survey of 1,050 shoppers.","source":"Attentive, 2026"},"fix":"Trains a multi-class model on what customers bought after what, and names the most likely next product for each one.","addsToEveryRow":"The most likely next product per customer, tested on how often its pick matched what happened.","workedExample":{"title":"The same campaign, pointed at the right product","saving":"$54,000","savingIs":"more revenue from the same campaigns, a year","working":"12 campaigns to 50,000 customers bring 30,000 orders from a one-size offer. 3% more from offers matched to each customer is 900 more orders, worth $54,000.","inputs":[{"id":"total","label":"Customers in one campaign","value":50000,"min":100,"max":500000000,"unit":null},{"id":"rate","label":"Who buy from a one-size offer","value":5,"min":0.5,"max":90,"unit":"%"},{"id":"campaigns","label":"Campaigns a year","value":12,"min":1,"max":365,"unit":null},{"id":"lift","label":"More sales from a matched offer","value":3,"min":0,"max":50,"unit":"%"},{"id":"value","label":"Value of one order","value":60,"min":1,"max":10000000,"unit":"$"}],"assumption":"The 3% is a starting figure, well below what published tests report: at Allegro, a new recommendation model lifted attributed sales by 21.25% in the web cart and 15.73% in the app.","isCustomerResult":false}},{"id":"demand","name":"Demand","internalClient":"Operations","question":"How much of each product will we need?","pain":{"figure":"44%","text":"of operators stock out at least once a month, and 74.3% run inventory on spreadsheets as their sole or primary tool, in a survey of 400 US operators.","source":"inFlow, 2026"},"fix":"Trains a forecaster on your sales history, product by product, and backtests it on recent months it was not shown. A quality gate checks the backtest and is built to refuse a forecast that fails it.","addsToEveryRow":"Units per product and period, backtested on recent months.","workedExample":{"title":"Less forecast error, less stock held just in case","saving":"$300,000","savingIs":"in cash released once, then $30,000 a year in holding cost","working":"By the textbook rule, safety stock is sized in proportion to forecast error. A 15% smaller error means 15% less safety stock: $300,000 released once, and $30,000 a year no longer spent holding it, at 10% a year.","inputs":[{"id":"stock","label":"Safety stock you hold","value":2000000,"min":1000,"max":10000000000,"unit":"$"},{"id":"cut","label":"Cut in forecast error","value":15,"min":0,"max":80,"unit":"%"},{"id":"carry","label":"Cost of holding stock, a year","value":10,"min":1,"max":60,"unit":"%"}],"assumption":"The 15% is a starting figure. Replace it with what your own backtest shows.","isCustomerResult":false}},{"id":"spend","name":"Spend","internalClient":"Finance","question":"What will each account spend next?","pain":{"figure":"18%","text":"of organisations can produce a forecast within two working days, and 44% can plan with high confidence for no more than three months.","source":"FP&A Trends, 2026"},"fix":"Builds a forecast for every account from its own history and backtests it on the most recent months. Where it cannot beat a simple baseline for an account, the baseline is used.","addsToEveryRow":"A forecast per account, month by month. The simple baseline where the model is no better.","workedExample":{"title":"A forecast for every account, every round","saving":"1,000 hours","savingIs":"of analyst time returned, a year","working":"500 accounts × 10 minutes × 12 rounds = 1,000 hours by hand. At ten minutes a run, 12 rounds take about 2 hours.","inputs":[{"id":"accounts","label":"Accounts to forecast","value":500,"min":1,"max":10000000,"unit":null},{"id":"minutes","label":"Minutes per account, by hand","value":10,"min":1,"max":600,"unit":"min"},{"id":"rounds","label":"Forecast rounds a year","value":12,"min":1,"max":365,"unit":null}],"assumption":"The minutes are your own estimate.","isCustomerResult":false}}]}