<\/div>\n <\/a>\n <\/div>\n\n \t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t\t

\n\t\t

\n\t\t\t\t

\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t $\"Red-Dot\"$ \t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t $\"Red$ \t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t\t

Modelling and testing <\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
AI modeling and testing is subject to robust AI governance processes and controls. Runtime algorithms are subject to strict SLAs and interaction models are refreshed nightly to ensure the best possible caller\/agent pairings<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Runtime algorithms<\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Runtime algorithms subject to strict SLAs and failsafe controls<\/p>
Pairing decisions made in < 50ms<\/li>
Preexisting client call center operational constraints are strictly adhered to<\/li><\/ul>
Callers are being paired with the existing set of agents<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Interaction model<\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Historical data is used to determine the expected payoff of different caller and agent pairings<\/p>
Interaction models and associated variables dynamically evolve to ensure best possible pairings<\/p>
Our models are retrained and refreshed nightly<\/li><\/ul>
OFF sample allows for analysis and improvements to the pairing models<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Validation <\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Gated process to validate across all stages<\/p>
Pre-deployment, deployment, and post-deployment<\/li><\/ul>
In the deployment phase Afiniti performs a cutover to Monitor Mode and validates every operational aspect before switching ON the solution<\/p>
Periodic testing conducted to ensure models and data used don\u2019t introduce bias (in conjunction with our clients, since Afiniti does not access\/process customer protected data).<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Sample algorithms<\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Utilization control<\/li>
Wait time thresholds<\/li>
Complex incumbent routing<\/li>
Call flow adaptively estimated<\/li>
Agent availability adaptively estimated<\/li><\/ul>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Sample models<\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Expectation maximization (EM)
clustering<\/li>
Maximum likelihood<\/li>
Bayesian models (including hierarchical)<\/li>
Generalized models (linear and non-linear)<\/li><\/ul>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Validation techniques<\/h4>\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
Control for over-fitting<\/li>
Matching strategy validation<\/li>
Comparative advantage validation<\/li>
Continuous ON\/OFF benchmarking<\/li><\/ul>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
Afiniti Risk Mitigation by Deployment Stage<\/h3>\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n
\n \n <\/div>\n\n
\n
\n
\n
\n
\n
01<\/div>\n
\n
Pre-deployment<\/h3>\n \n \n \n <\/svg>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n
Global data risk categorization<\/h6>\n
(low, medium, and high) to deter bias and discrimination (including known proxies)<\/p>\n
Process controls to remediate medium and high risk data elements in the collection and organization stages of Afiniti\u2019s data supply chain<\/p>\n <\/li>\n
Data elements documented<\/h6>\n
to include purpose, risk level, location, etc.<\/p>\n <\/li>\n
Dual data architecture<\/h6>\n
supporting Afiniti\u2019s purpose: Discovery and Consumption<\/p>\n <\/li>\n
Organizational controls<\/h6>\n
to implement in each stage of the data supply chain when processing, labeling and preparing data for consumption<\/p>\n <\/li>\n <\/ul>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n
\n
\n
02<\/div>\n
\n
Deployment<\/h3>\n \n \n \n <\/svg>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n
Phased approach<\/h6>\n
Monitor Mode > FIFO > Cutover to validate before switching ON the solution <\/p>\n <\/li>\n
Priority signals<\/h6>\n
(more than 50) are part of our deployments i.e., signals for client assurance including agent utilization and caller waiting times<\/p>\n <\/li>\n
Continuous benchmarking<\/h6>\n
ON and OFF benchmarking to review impact of Afiniti across customer segments available from structured data<\/p>\n <\/li>\n
Configurable<\/h6>\n
to constrain potential bias based on metrics<\/p>\n <\/li>\n
Model re-training<\/h6>\n
system for mitigating any potential bias<\/p>\n <\/li>\n <\/ul>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n
\n
\n
03<\/div>\n
\n
Post-deployment<\/h3>\n \n \n \n <\/svg>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n
Access to variables<\/h6>\n
(and learning ability for model performance monitoring and logging over time<\/p>\n <\/li>\n
Signals and controls<\/h6>\n
in place to raise alarms<\/p>\n <\/li>\n
Model historical views<\/h6>\n
(production uptime, revenue impact, operational impact)<\/p>\n <\/li>\n
Afiniti can<\/h6>\n
temporarily switch back to Monitor mode as needed for risk mitigation purposes<\/p>\n <\/li>\n <\/ul>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n<\/div>\n\n