# Tandem (10dem) > 10dem is an applied AI consultancy for mid-market consumer tech. It builds AI solutions one business function at a time and stays until they show up in the P&L. Two solutions are live. Marketing: SEO and AEO, getting companies ranked on Google and cited by AI answer engines. Analytics: decision models that are deterministic, so every number is reproducible and traceable. ## Key pages - [Marketing: SEO & AEO](https://www.10dem.xyz/marketing-seo-aeo): Getting found on Google and cited in AI answers - [Analytics: decision models](https://www.10dem.xyz/decision-analytics): Decisions on deterministic, auditable models - [AI solutions](https://www.10dem.xyz/services): The live solutions and how we engage - [How we work](https://www.10dem.xyz/how-we-work): Our 90-day delivery model - [Insights](https://www.10dem.xyz/blog): Essays and frameworks on applied AI - [Contact](https://www.10dem.xyz/contact): Book a free 30-minute audit ## AI solutions (live) - [Marketing: SEO and AEO](https://www.10dem.xyz/marketing-seo-aeo): Getting consumer-tech companies ranked on Google and cited by AI answer engines (ChatGPT, Perplexity, AI Overviews) - [Analytics: deterministic decision models](https://www.10dem.xyz/decision-analytics): Forecasting, pricing, attribution and LTV on models where every number is reproducible and traceable ## Other services - [AI adoption consulting](https://www.10dem.xyz/ai-adoption-consultant): Fractional, outcome-measured AI adoption for mid-market operators - [Fractional AI team](https://www.10dem.xyz/fractional-ai-team): A senior AI team on tap, without full-time hires - [AI training for leadership](https://www.10dem.xyz/training/leadership): Executive fluency on where AI moves the P&L - [AI training for teams](https://www.10dem.xyz/training/teams): Hands-on workflows training for operating teams ## AI use-case guides - [AI Personalization for DTC & Ecommerce Brands](https://www.10dem.xyz/ai-for/ai-personalization-for-dtc-ecommerce-brands): How mid-market DTC and ecommerce brands use AI personalization to lift AOV and repeat purchase, where the value really sits, and why most projects stall on workflow, not the model. - [AI Customer Support Automation for Subscription Apps](https://www.10dem.xyz/ai-for/ai-customer-support-automation-for-subscription-apps): How subscription apps automate tier-1 support with AI to cut resolution time and cost, where automation pays off, and why escalation design and team adoption decide whether it sticks. - [AI Churn Prediction for Subscription Apps](https://www.10dem.xyz/ai-for/ai-churn-prediction-for-subscription-apps): How subscription apps use AI churn prediction to retain revenue, why a churn score alone changes nothing, and how the intervention workflow is what actually moves retention. - [AI Demand Forecasting for DTC Brands](https://www.10dem.xyz/ai-for/ai-demand-forecasting-for-dtc-brands): How DTC brands use AI demand forecasting to cut stockouts and dead inventory, why the forecast is the easy part, and how the planning process decides whether it pays off. - [AI Recommendation Engines for Marketplaces](https://www.10dem.xyz/ai-for/ai-recommendation-engines-for-marketplaces): How marketplaces use AI recommendation engines to lift GMV and liquidity, how to handle cold start, and why ranking objectives and merchandiser trust decide the outcome. - [AI Marketing Automation for DTC Brands](https://www.10dem.xyz/ai-for/ai-marketing-automation-for-dtc-brands): How DTC brands use AI marketing automation to lift repeat revenue and cut creative cycle time, which flows actually pay, and why the operating cadence matters more than the tool. - [AI Pricing Optimization for Ecommerce Brands](https://www.10dem.xyz/ai-for/ai-pricing-optimization-for-ecommerce-brands): How ecommerce brands use AI pricing optimization to protect margin and time markdowns, where elasticity models genuinely help, and why pricing governance decides the outcome. - [AI Content Generation for Ecommerce Catalogs](https://www.10dem.xyz/ai-for/ai-content-generation-for-ecommerce-catalogs): How ecommerce teams use AI to generate product descriptions, attributes, and SEO content at catalog scale, what quality control actually requires, and why structure beats prose. - [AI Fraud Detection for Marketplaces](https://www.10dem.xyz/ai-for/ai-fraud-detection-for-marketplaces): How marketplaces use AI fraud detection to cut chargebacks and seller abuse without strangling growth, and why the false-positive budget and ops workflow matter more than the model. - [AI Onboarding Personalization for Subscription Apps](https://www.10dem.xyz/ai-for/ai-onboarding-personalization-for-subscription-apps): How subscription apps use AI to personalize onboarding and lift activation and trial conversion, why the first session decides retention, and why flow redesign matters more than the model. - [AI LTV Prediction for Subscription Apps](https://www.10dem.xyz/ai-for/ai-ltv-prediction-for-subscription-apps): How subscription apps use AI to predict customer lifetime value early, and why the payoff comes from wiring predictions into acquisition bidding and lifecycle decisions, not from the model. - [AI Review & Feedback Analysis for Consumer Apps](https://www.10dem.xyz/ai-for/ai-review-and-feedback-analysis-for-consumer-apps): How consumer apps use AI to mine app store reviews, support tickets, and surveys for product decisions, and why the routing of insights into roadmaps matters more than the analysis. - [AI Search Relevance for Marketplaces](https://www.10dem.xyz/ai-for/ai-search-relevance-for-marketplaces): How marketplaces use AI to improve search relevance and lift conversion, why zero-result and mis-ranked queries quietly drain GMV, and why the tuning process matters more than the algorithm. - [AI Inventory Optimization for DTC Brands](https://www.10dem.xyz/ai-for/ai-inventory-optimization-for-dtc-brands): How DTC brands use AI inventory optimization to free working capital and cut stockouts, what it adds beyond forecasting, and why the buying process decides whether it pays off. - [AI Customer Segmentation for Consumer Brands](https://www.10dem.xyz/ai-for/ai-customer-segmentation-for-consumer-brands): How consumer brands use AI customer segmentation to target spend and lifecycle messaging by behavior and value, and why segments only pay off when they are wired into activation. ## Case studies - [Procurement across 8 collection centers. One 30-minute refresh.](https://www.10dem.xyz/work/dairy-procurement-dashboard): How a scaling dairy D2C brand went from a 2–3 day, error-prone procurement refresh across 8 collection centers to a 30-minute, farmer-level view of the whole network. - [Production shortfall and inventory cost, seen 12 weeks ahead](https://www.10dem.xyz/work/ev-production-inventory-planning): How a large EV manufacturer replaced a weekly ops-finance reconciliation scramble with real-time planning: daily vehicle plan vs. parts shortfall, supplier capacity, and working-capital exposure seen 6 weeks earlier. - [Sales analytics live from the ERP, and 3x revenue since](https://www.10dem.xyz/work/fmcg-sales-analytics): How a multi-region FMCG manufacturer replaced a 2-day manual Excel with live Power BI analytics across 10 locations, restructured for growth, and grew revenue 3x since 2021. - [Month-end close: 2–3 days to 1 hour](https://www.10dem.xyz/work/saas-month-end-close): How four SaaS companies with $2M combined MRR cut month-end close from 2–3 days to 1 hour with ETL automation, cutting reconciliation errors 80% and redirecting finance time to variance analysis. - [Demand planning that runs ahead of demand](https://www.10dem.xyz/work/food-demand-planning): How a food & baking company replaced 3 weeks of manual demand compilation with a data-driven model: 25% better forecast accuracy, 28% less excess inventory, and margins recovered from waste. - [45,000 SKUs classified. Inventory stopped eating cash.](https://www.10dem.xyz/work/retail-inventory-optimization): How a premium multi-outlet retailer classified 45,000 SKUs, redefined roles and KPIs, and closed ERP process gaps to cut tied-up inventory and lift inventory turnover. - [From 10 retailers to a tracked network, and ₹80–220 Cr growth in 18 months](https://www.10dem.xyz/work/construction-materials-distribution): How a construction materials manufacturer went from ~10 retailers and zero secondary-sales visibility to a tracked, incentivized distribution network, delivering ₹80–220 Cr growth within 1.5 years. - [Manpower norms built from the factory floor, not the org chart](https://www.10dem.xyz/work/paints-manpower-optimization): How a paint production unit cut man-hour costs with manpower norms built from real task data across 80% of batches, validated against shifts and edge cases, and earning an internal cost-optimization award. - [Turning a churn score into a retention system](https://www.10dem.xyz/work/subscription-churn-retention): An illustrative engagement: how a subscription app turned an unused churn score into a retention system that its team runs every day. Representative of our approach. - [Personalization that merchandisers actually run](https://www.10dem.xyz/work/dtc-personalization-aov): An illustrative engagement: how a DTC brand turned an unused recommender into a personalization system its merchandisers run weekly. Representative of our approach. ## Articles - [Why the Forward Deployed Engineering Model Works for Mid-Market AI](https://www.10dem.xyz/blog/forward-deployed-engineering-model-mid-market-ai): Mid-market companies do not fail at AI because they picked the wrong model. They fail in the gap between a plan and production. Forward deployed engineering exists to close that gap. - [Forward Deployed Engineer vs Solutions Engineer vs Consultant](https://www.10dem.xyz/blog/forward-deployed-engineer-vs-solutions-engineer): Three client-facing technical roles, three completely different accountabilities. Hiring the wrong one is the most common and most expensive staffing mistake in AI delivery. - [How to Hire a Forward Deployed Engineer (Skills, Interview, Comp)](https://www.10dem.xyz/blog/how-to-hire-a-forward-deployed-engineer): The forward deployed engineer is the hardest technical seat to hire for, because the thing that makes someone good at it is invisible to a standard interview loop. Here is what to screen for instead. - [What Is a Forward Deployed Engineer? (And Why AI Work Needs One)](https://www.10dem.xyz/blog/what-is-a-forward-deployed-engineer): A forward deployed engineer is an engineer who works inside the customer's environment instead of behind a roadmap. In AI work the model is rarely the hard part, so the person closest to the messy reality is the one who makes it pay off. - [Your AI Can Now Write Tests. Here's Why Your QA Math Just Changed.](https://www.10dem.xyz/blog/your-ai-can-now-write-tests-heres-why-your-qa-math-just-changed): Microsoft's new AI test generator rewrites the economics of QA for mid-market teams. Unit test writing costs drop 70%, coverage gaps close in weeks instead of quarters, and your QA spend shifts from labor-intensive test writing to strategic test quality. - [How to Calculate AI ROI: A Practical Framework for Mid-Market Companies](https://www.10dem.xyz/blog/how-to-calculate-ai-roi-a-practical-framework-for-mid-market-companies): Most mid-market companies struggle to measure AI ROI because they confuse cost savings with value creation. Here's the framework ops leaders use to move from speculation to numbers. - [The Model Switching Trap: When Your Cheap AI Strategy Costs You More](https://www.10dem.xyz/blog/the-model-switching-trap-when-your-cheap-ai-strategy-costs-you-more): Switching to a cheaper AI model often costs more. Learn the framework for routing tasks to different models based on economics, not just price-per-token. - [When to Build Your Own AI: The Hidden Costs of Off-the-Shelf Models](https://www.10dem.xyz/blog/when-to-build-your-own-ai-the-hidden-costs-of-off-the-shelf-models): Most mid-market companies buy off-the-shelf AI models first. But there are situations where building pays off. Here's how to decide. - [How to Pick Your First AI Project (Without Wasting Six Months)](https://www.10dem.xyz/blog/how-to-pick-your-first-ai-project-without-wasting-six-months): Most companies pick their first AI project backwards. They chase the sexiest idea or the one that got pitch deck attention, then spend months discovering it was unsolvable or unprofitable. Here is how operators actually pick winners. - [AI ROI Under Scrutiny: How to Prove Value When the Money Gets Tighter](https://www.10dem.xyz/blog/ai-roi-under-scrutiny-how-to-prove-value-when-the-money-gets-tighter): As capital dries up and AI spending slows, mid-market operators need hard numbers to justify continued investment. Here's how to measure real ROI, not hype. - [Running AI in Production Without Blowing Up Your Stack: Deployment Patterns for Mid-Market](https://www.10dem.xyz/blog/running-ai-in-production-without-blowing-up-your-stack-deployment-patterns-for-mid-market): Deploy AI models safely using shadow mode, canary rollouts, and circuit breakers. Real patterns for mid-market teams to avoid production disasters. - [The Data Quality Cost: Why Your AI Training Data Matters More Than Model Choice](https://www.10dem.xyz/blog/the-data-quality-cost-why-your-ai-training-data-matters-more-than-model-choice): AI model performance depends more on training data quality than on which model you pick. Operators who invest in structured data collection and validation before training reduce downstream costs by 40-60% and cut debugging time from weeks to days. - [The Hidden Ops Tax: Why Your AI Headcount Never Matches Your Model](https://www.10dem.xyz/blog/the-hidden-ops-tax-why-your-ai-headcount-never-matches-your-model): You budgeted for an ML engineer and a data analyst. Then your AI hit production. Now you need three more people just to keep it from breaking. Here's why operators consistently underestimate team size for production AI. - [The Inference Cost Trap: Why Your AI Budget Is About to Blow Up](https://www.10dem.xyz/blog/the-inference-cost-trap-why-your-ai-budget-is-about-to-blow-up): Your pilot cost $500/month. Scale it to production with always-on agents, and you are looking at $8,000/month. Here's how to predict and control the explosion before month two. - [Why Your AI Spend Is About to Get Worse (and How to Model It)](https://www.10dem.xyz/blog/why-your-ai-spend-is-about-to-get-worse-and-how-to-model-it): Most mid-market companies flying on AI pilot budgets are about to hit a wall. Here's how to build a cost model before you do. - [Fine-Tuning vs. Prompt Engineering: When the $50K Investment Actually Pays Off](https://www.10dem.xyz/blog/fine-tuning-vs-prompt-engineering-when-the-50k-investment-actually-pays-off): Your team wants to fine-tune a model for better performance. The estimated cost: $30-50K in data prep, training, and infrastructure. But you could get 80% of the benefit from better prompts and evals for $5K. Here's how to know which path actually matters. - [Hire AI Talent or Buy Services? A Mid-Market Decision Framework](https://www.10dem.xyz/blog/hire-ai-talent-or-buy-services-a-mid-market-decision-framework): A $50M revenue company faces a choice: hire a 3-person ML engineering team ($400K+ annually) or use a managed AI service ($100-200K annually). The wrong choice costs you either excess overhead or months of lost velocity. Here's how to decide. - [India's DPDP Act and Your AI: What Operators Need to Know Now](https://www.10dem.xyz/blog/indias-dpdp-act-and-your-ai-what-operators-need-to-know-now): India's Digital Personal Data Protection Act forces a hard choice for mid-market AI operators: redesign your data pipelines for consent and purpose limitation, or accept severe constraints on AI training and personalization. The rules are vague, enforcement is phased, and time is running out. - [Why Your AI Pilot Will Fail (And When to Go Phased Instead of Big-Bang)](https://www.10dem.xyz/blog/why-your-ai-pilot-will-fail-and-when-to-go-phased-instead-of-big-bang): Your AI initiative has a 60% chance of stalling in pilot. The reason isn't the model or the data. It's usually the deployment decision you made before you wrote a single line of code. - [Is Your AI-Written Code Secure? The Vibe-Coding Security Gap](https://www.10dem.xyz/blog/is-your-ai-written-code-secure-the-vibe-coding-security-gap): AI does not write insecure code randomly. It fails hardest on exactly the things that matter most: authentication, access control, and your cloud config. The average score hides the danger. Here is what the data actually shows and what to do about it. ## Full content - [llms-full.txt](https://www.10dem.xyz/llms-full.txt): Complete text of the use-case guides and case studies in one document ## Profiles - https://www.instagram.com/10dem.xyz/ - https://x.com/_10dem - https://www.linkedin.com/company/136014120/ - https://www.linkedin.com/in/7ankurg/ - https://www.linkedin.com/in/gaurav-bhushan-sharma/ ## Contact hello@10dem.xyz ## Usage Content may be quoted and cited with attribution to 10dem (10dem.xyz).