Machine Learning & LLM Engineer Predictive Intelligence

Level : Mid-Senior to Senior

Function : Machine Learning & AI Engineering

Employment Type : Full-Time

Location : Remote

Working Hours : 2:30 PM - 11:30 PM IST

Company Overview :

At Codvo, software and people transformations go hand-in-hand. We are a global empathy-led technology services company. Product innovation and mature software engineering are part of our core DNA.

Respect, Fairness, Growth, Agility, and Inclusiveness are the core values that we aspire to live by each day. We continue to expand our digital strategy, design, architecture, and product management capabilities to offer expertise, outside-the-box thinking, and measurable results.

Role Purpose :

Support a machine learning initiative to transform an existing knowledge/competitor-oriented agent into a product winner prediction agent.

Key Responsibilities :

- Work on Machine Learning and LLM-based use cases.

- Transform an existing agent into a system that predicts whether a product will be a commercial winner or non-winner.

- Build workflows where users input product specifications such as color, fabric, sleeve type, and other attributes.

- Develop predictive models tied to assortment and product performance data.

About the Role :

We are looking for a talented Machine Learning & LLM Engineer to work on a high-impact initiative at the intersection of traditional predictive modeling and modern Large Language Models (LLMs).

Your primary mission is to transform an existing knowledge and competitor intelligence agent into a product winner prediction enginea system capable of assessing product specifications such as color, fabric, sleeve type, category, and price point to predict whether a product is likely to be a commercial winner or non-winner.

This role requires strong expertise in both classical Machine Learning and hands-on LLM engineering. You will work with real product and assortment data, design prediction workflows, and build intelligent pipelines that turn product attributes into actionable commercial insights.

What You Will Do :

- Lead the transformation of an existing knowledge/competitor-oriented agent into a product winner prediction agent by redesigning its core intelligence layer from retrieval and lookup to predictive scoring.

- Design and build ML prediction pipelines that take structured product inputs (color, fabric, sleeve type, category, price point, etc.) and generate winner/non-winner classifications with confidence scores.

- Develop and optimize LLM-integrated workflows where natural language product descriptions, buyer briefs, or specification sheets are parsed, enriched, and fed into predictive models.

- Build user-facing workflows that allow business users to enter product specifications through structured or conversational interfaces and receive ranked predictions with explanations.

- Work with assortment and product performance datasets to build, validate, and continuously improve supervised and semi-supervised predictive models.

- Engineer feature extraction pipelines from product attributes, including :

1. Color

2. Fabric

3. Construction

4. Seasonal trends

5. Historical sell-through rates

6. Competitor signals

- Collaborate with product and data teams to define winner/non-winner labeling strategies and identify business metrics such as :

1. Sell-through rate

2. Margin

3. Reorder rate

- Evaluate, benchmark, and improve model performance through offline evaluation frameworks and production feedback loops.

- Document model architecture, data lineage, and prediction logic to support governance, explainability, and stakeholder trust.

Required Skills & Experience :

1. Machine Learning :

- Strong hands-on ML background.

- Classification and regression modeling.

- Ensemble methods :

a. XGBoost

b. LightGBM

c. Random Forest

- Feature engineering.

- Model evaluation and production deployment.

2. LLM Engineering :

- Practical experience integrating LLMs into production workflows.

- Prompt engineering.

- Function/Tool calling.

- RAG (Retrieval-Augmented Generation) pipelines.

- Output parsing.

- LLM evaluation.

3. Predictive Product / Outcome Modeling :

- Experience building models that predict commercial or product outcomes from structured attribute data.

- Retail, fashion, FMCG, or assortment planning experience is highly desirable.

4. Data & Feature Engineering:

- Strong Python skills.

- Experience with :

a. Pandas

b. NumPy

c. Scikit-learn

- Ability to clean, wrangle, and engineer features from product catalog and transactional datasets.

5. ML Pipeline & Deployment :

- Experience building and deploying end-to-end ML pipelines including :

a. Training

b. Evaluation

c. Versioning

d. Inference serving

Preferred Skills & Experience :

1. Retail / Fashion / Assortment Data :

- Product assortment data, merchandising systems, PLM data, demand forecasting, and consumer goods analytics.

2. LLM Frameworks :

- LangChain, LlamaIndex, Semantic Kernel, and similar orchestration frameworks.

3. Embedding & Similarity Models :

- Text embeddings, multimodal embeddings, similarity search, and product clustering.

4. MLOps & Model Governance :

- MLflow, Weights & Biases, experiment tracking, model registry, and performance monitoring.

5. Cloud ML Platforms :

- Azure ML, AWS SageMaker, and Google Vertex AI.

6. Agentic / Multi-Step Workflows :

- Experience building agentic pipelines where LLMs orchestrate tool calls, data lookups, and model inference workflows.

What We're Looking For :

- 4- 8 years of experience in Machine Learning and/or Applied AI Engineering.

- At least 2 years of hands-on experience working with LLMs in production or near-production environments.

- Strong quantitative foundation with expertise in :

1. Classification models

2. Probability calibration

3. Evaluation metrics

4. AUC

5. F1 Score

6. Precision/Recall trade-offs

- Comfortable working with both structured tabular data (product attributes, sales history) and unstructured text (product descriptions, buyer notes, trend reports).

- Ability to balance model sophistication with delivery speed.

- Strong collaboration and stakeholder management skills.

- Genuine curiosity about understanding what drives commercial product success.

Nice to Have :

- Experience with multimodal models that combine product images and structured product attributes.

- Familiarity with active learning or human-in-the-loop labeling workflows.

- Exposure to A/B testing frameworks for validating prediction outcomes.

- Prior experience transforming a rule-based or retrieval-based system into a machine-learning-powered solution.