Merciv - Senior ML/AI Engineer, Agentic Intelligence
At a glance
Senior AI Engineering role at deCircle. New York · full-time · no salary stated · we estimate $178,750–$239,488.
deCircle has not published a salary for this role. The figure below is our estimate, not theirs.
aimlroles summary, based on the employer's posting.
What you'll do
- Build and launch models for demand forecasts, time-series predictions, consumer sentiment and anomaly detection
- Develop agent-based systems that reason over varied data sources and carry out actions independently
- Run ML workflows from data preparation and feature design through training, evaluation and production release
- Create RAG and LLM components for natural-language interfaces and automated workflows
- Track deployed model performance, retrain when needed and improve reliability with backend engineers
What you bring
- At least five years applying ML or AI, including models already running in production
- A graduate degree in computer science, machine learning or statistics, or equivalent practical experience
- Strong Python skills and practical work with PyTorch, TensorFlow or scikit-learn
- Experience with predictive statistics, time-series forecasting, agentic systems, NLP, LLMs and RAG
Who this fits
This suits a senior applied ML engineer who wants to take models from development through production and maintain their performance. The role is onsite in New York. You’ll work as an individual contributor alongside backend engineers; former technical founders are encouraged to apply.
From the employer
Merciv is building the intelligence layer for enterprise commerce. We connect an organization's entire data landscape — internal systems, social signals, industry reports, consumer behavior — and surface the insights and automated workflows that used to take analysts weeks.
We raised $14M in seed funding, spent two years in stealth building the right thing, and we're now launching publicly. Our platform already drives 8-figure gross margin improvements for some of the world's largest retailers.
We're a lean, high-ownership team. If you want to see your work matter immediately, this is it.
The Role
You'll build and deploy the intelligent systems at the core of Merciv. Our platform doesn't just surface insights — it reasons, forecasts, and acts autonomously across complex enterprise data landscapes. You'll develop the models and agentic architectures that power demand forecasting, consumer intelligence, competitive analysis, and autonomous decision-making for the world's largest retailers.
This is applied AI at its most impactful. You'll work at the intersection of cutting-edge research and real enterprise deployment, building systems that generate tens of millions in value for clients.
What You'll Do
Design, build, and deploy ML models for demand forecasting, time series prediction, consumer sentiment analysis, and anomaly detection at enterprise scale
Develop and iterate on Merciv's agentic AI architecture — building systems that reason across heterogeneous data sources and take autonomous action
Build and maintain robust ML pipelines: data preprocessing, feature engineering, model training, evaluation, and production deployment
Architect RAG systems and LLM integrations that power Merciv's natural language interfaces and autonomous workflows
Collaborate with backend engineers to ensure models are production-grade — optimized for latency, reliability, and scale
Own model performance end-to-end: monitoring, retraining, and continuous improvement in production
Stay at the frontier of AI research and bring relevant innovations into the platform
You have 5+ years of applied ML/AI experience with models deployed in production
You have an M.S. or Ph.D. in CS, Machine Learning, Statistics, or equivalent practical experience
You're deeply proficient in Python with hands-on experience in PyTorch, TensorFlow, or scikit-learn
You have a strong background in statistical analysis, predictive modeling, and time series forecasting
You've worked on agentic AI systems, multi-agent orchestration, NLP, LLMs, and RAG architectures
You care about model interpretability and building systems enterprise users can actually trust
Former technical founders are encouraged to apply — we care more about what you've built than how long you've been building