Build, deploy & govern Agentic AI apps with SAS Viya - free on #LearnSAS ✔️Get hands-on with LLMs, ModelOps & DecisionOps: 2.sas.com/6013B8FyC3

意思決定をskill化してDecisionOpsしたい

SmoothPath Origin The input was two words: silken and quandary. Silken suggests distributing tension so nothing snaps. Quandary means a knot of conflicting constraints. Put together, the core idea was to soften hard trade-offs so decisions can move without breaking. That led us to model choices as a fabric rather than a tree or a vote — with soft constraints and graded penalties instead of binary gates. From there, we defined a new category: DecisionOps, software for continuous, explainable decision making. Name SmoothPath is the recommended product and company name. It’s short, memorable, and works in both developer and business contexts. Path signals a clear direction that teams can follow, while Smooth communicates ease and frictionless decision-making. Concept SmoothPath turns hard team trade-offs into a live decision fabric. Teams import backlog, budgets, or goals, set criteria, and drag weights. The system generates Pareto-efficient option sets, highlights where hard constraints bind, and explains the trade-offs in plain language. Over time it learns an organization’s Decision DNA by mapping inputs → choices → outcomes. Target Initial focus is 50–500-person SaaS and fintech orgs where product and engineering teams share limited capacity. The beachhead wedge is roadmap arbitration for PM and Eng leads working in Linear or Jira with GitHub. MVP Backend: Laravel API with PostgreSQL Frontend: Next.js application Schema: projects, criteria, options, scores, constraints, runs, decisions, integrations API: endpoints for creating projects, adding criteria and options, posting constraints, running solver, retrieving results Solver: elastic multi-objective scoring with soft penalties; greedy seed local swap simulated annealing; maintains a Pareto front for value, cost, and risk UI: sliders for weights, Pareto chips, tear map, and rationale panel Integrations: Linear or Jira import, Slack command, GitHub labels Business Model Free tier: one project and basic import Pro tier: $29/mo with unlimited criteria and Slack integration Team tier: $99/mo with SSO and private templates Business tier: $499/mo with on-prem solver and SOC2 compliance Enterprise: custom pricing and support Moat Private Decision DNA corpus that improves recommendations over time Deep in-flow presence inside PRs, tickets, and chat — not a separate destination Template marketplace for common decision fabrics Solver tuned by each organization’s historical deltas Build Timeline 0–30 days: CSV import MVP, Linear and Slack integrations, first design partners 31–60 days: outcome tracking, Pareto archive, rationale generation, first case study 61–90 days: constraint language v1, GitHub Action, SSO, two enterprise pilots

How can #Streamlit-based UIs help accelerate stakeholder buy-in for decision models? Join us Thursday as the @AMPLopt and Nextmv teams explore just that with a live demo of a Streamlit app for a facility location decision model: hubs.la/Q02x53cG0 #ampl #decisionOps #orms

Treating Decision Automation like a typical software dev project is expensive and time consuming. There is a better way - to establish a lean and self-sufficient DecisionOps team. ▶️ flexrule.com/links/p1ki #decisionautomation #decisionintelligence #businessrules

Can you deploy and test a custom decision model with Nextmv? Most definitely! Check out how in our docs: nextmv.io/docs/platform/depl… You can also see recent examples with Java OR-Tools at nextmv.io/videos #python #java #golang #decisionscience #orms #AI #DecisionOps

It's great to be at INFORMS Analytics 2024! Come visit us at booth 205 to chat DecisionOps, CI/CD, and testing for custom decision models featuring integrations for OR-Tools, Gurobi, AMPL, Pyomo, HiGHS, and more! #analytics2024 #orms #informs #decisionscience #cicd #python

How do you evaluate different #Gurobipy model formulations? Learn how with #DecisionOps tooling! Join our getting started techtalk on Wednesday, April 10: hubs.la/Q02sjlJ60 #orms #logistics #gurobi #decisionscience

How do you test if you should use map data over Haversine for a vehicle routing problem? We use batch experiments with Nextmv to find out for an example use case: hubs.la/Q02rGj0b0 #orms #DecisionOps #routeoptimization #logistics #fleetoptimization #vehiclerouting

Create an operational OR-Tools decision app in minutes. Separate your operational data from your model, automate it as a decision service, and interact with it using simple API calls. hubs.la/Q02r3h-C0 #decisionscience #orms #logistics #decisionops #ortools #python #java

🐇🐈‍⬛🎉Do you build AMPL decision models? Check out the latest integration that pairs all the goodness of @AMPLopt with Nextmv’s DecisionOps platform: hubs.la/Q02pZ90k0 #decisionscience #orms #decisionOps #ampl #datascience #operationsresearch #cicd

Going to @2024_analytics? We'll be there hosting a workshop and staffing a booth! We’re looking forward to seeing everyone! hubs.la/Q02pRrtR0 #informs #orms #decisionscience #2024analytics #decisionOps

New release! 🎉🎉🎉 The Nextmv Gurobi integration accelerates how you run, test, and deploy decision models running @gurobi with Nexmv’s DecisionOps platform: hubs.la/Q02nyvvK0 #orms #decisionops #decisionscience #gurobi #optimization #datascience

📣 New release! We’re proud to offer switchback testing on the Nextmv platform for production testing two models with operational impacts. Get a primer on switchback testing and watch a demo of performing and analyzing an experiment >> hubs.la/Q02hr97G0 #decisionops #orms

👋Hello to the Nextmv Pyomo integration: Build, test, and deploy Python-based Pyomo decision models even faster. Get the rundown (with a slick video demo): hubs.ly/Q02d3Scv0 #pyomo #orms #python #decisionOps #optimization #glpk #datascience

🔥 Autumn Product Release 🔈 6 highlights from our last 3 monthly product releases are below. Register to today's product launch webinar to learn more: causalens.com/resources/webi… You will get automatically get the recording in case you can't make it. ---- 1️⃣ Introducing the next generation of developer experience We have launched a new developer environment to help Data Scientists get started with Causal AI much faster. Our current Beta users love: - The MetaCell functionality, allowing them to use pre-built code templates for specific product features (e.g. Structural Causal Modelling) as well as use-case-specific workflows. - Sharing your apps with both Business Stakeholders and Data Science users at the click of a button! Plenty of exciting features are in the pipeline in order to offer the best experience to our data science users. 2️⃣ Gen AI 🤝 Causal AI Our new Lexicause package leverages LLMs to enhance Causal AI workflows, introducing prompt engineering for causal discovery, causal modelling and causal reasoning. At the same time, Causal AI can ground LLMs in reality and help mitigate hallucinations by passing causal knowledge into prompts. 3️⃣ decisionOps Metacells Too often, Data Science projects are seen as proof of concepts or laboratory experiments, usually because there is no clear way to attribute their impact on KPIs or measure ROI. Users can now orchestrate decision workflows, scale decision-making to meet their needs and measure and optimize the ROI of their decision-making with code templates We continue to help customers put trust back into their decision-making, and so there are now new code templates (MetaCells) available for: ✅ KPI Calculation ✅ KPI Visualisation ✅ Action - you can now define arbitrary code as a repeatable action. ✅ decisionStore for saving decisions. It takes just a single parameter to point it at your chosen folder to store the decision. 4️⃣ Decision Optimization We introduced the cAI-optimization package! This package uncovers the best interventions within structural causal models to maximize (or minimize) an objective function. 5️⃣ Time Series & Causal Discovery We've added two time-series-specific Causal Discovery algorithms. ✅ DYNOTEARS, which is a continuous-optimization causal discovery algorithm ✅ PCMCIPlus, which is a constraint-based causal discovery algorithm We keep expanding our causal discovery library to offer the most complete & advanced enterprise platform for Causal AI. 6️⃣ Open Sourcing In August, we open-sourced the core package of Dara, our app-building framework. Github: github.com/causalens/dara One of the most significant components included in the open sourcing is the Casual Graph Visualization Package, and the latest addition to the package is the Time Series Causal Graph, which helps Data Scientists construct temporally-aware graphs. Hope you enjoy the latest features. We will keep shipping in the meantime 🚀

"The term 'prescription' implicitly carries the meaning of causality." "Is [decision] optimization an iterative process over counterfactual scenarios?" 2 great industry panels with representatives from @AmericanExpress @eBay @Fidelity @Zalando @Microsoft @IBM - recording: youtube.com/watch?si=NXXhQGO… Proud to support the @causal_science 2023! Our platform, decisionOS, combines causal and prescriptive science in an enterprise-ready platform to help companies translate data and domain knowledge into better decision workflows in production. Our human-guided causal discovery framework combines the best of domain knowledge and algorithmic approaches to discover causal graphs: causalens.com/human-guided-c… Our structural causal model, causalNet, discovers optimal structural causal models that can both robustly predict and accurately estimate interventions and counterfactuals: causalens.com/causalnet-stat… Our decision optimization engine discovers the optimal decision, optimizing over counterfactual and interventional scenarios given constraints and cost functions. One approach is algorithmic recourse: causalens.com/algorithmic-re… Our decisionOps framework measures the effect of decisions: causalens.com/decisionops/

At look at an example CI/CD flow for improving on a vehicle routing (CVRP) decision model using OR-Tools, GitHub Actions, and Nextmv: hubs.ly/Q028kL1L0 #DecisionOps #orms #informs #ortools #routeoptimization #CICD #datascience

Great talk by Alexandre Trilla from @Alstom at the Causal AI conference 2023! "It turns out that predictive technology is quite limited in these segments... and causal inference is precisely the science that allows us to intervene, change it, and evaluate it. Predictive systems are focused on mapping inputs to outputs, and they will do whatever it takes to minimize the classification error or regression loss. This often renders them uninterpretable, which penalizes the confidence of our solutions. Causal systems are focused on modeling these variable interactions. This makes it clear what's going on under the hood and can also help us deliver solutions with a greater degree of confidence" Great motivation for "why Causal AI" 2 great use cases presented. Thank you for being with us, Alexandre 🙏 Our platform is addressing these challenges of traditional, predictive, machine learning: ✅ Our Human Guided Causal Discovery framework combines the best of human knowledge data-driven approaches to discover cause-effect relationships ✅ Our Structural Causal Model, causalNet, helps enterprises build inherently explainable models that can both robustly predict as well as estimate the effect of interventions & counterfactuals ✅ Our root cause analysis engine helps uncover the true root causes of inefficiencies and faults ✅ Our decisionOps framework measures the impact of decisions The talk: youtube.com/watch?v=p-JsUVcH… More on our platform: causalens.com/decision-os/

"Causal questions are in 1st order in tech firms." "Lots of good executives have a causal theory of reality in their heads. What they often want to see is... dashboards, metrics... these highly correlated data where the causal filtering is happening in a board room, in a meeting, in a discussion. The interesting challenge for this [Causal AI] community is now that we have these [Causal AI] tools [and] you don't have to do it intuitively, how do you reconcile this with the traditional mindset?" Great discussion with @AmitEcon and @pedrohcgs at the Causal AI conference 2023! youtu.be/D0AREV5mk6w?si=DZjI… Our human-guided causal discovery framework and causal model architectures help enterprises discover cause-effect relationships and quickly develop state-of-the-art causal models. Our decisionOps framework helps businesses measure the effect of their decisions on business KPIs. Thank you both for being with us! 🙏