Generative AI Platform | TeraSystemsAI Careers

Platform Overview

Safe and Controllable Generative AI

Our Generative AI Platform delivers state-of-the-art foundation models with enterprise-grade safety, alignment, and governance. Through constitutional AI, reinforcement learning from human feedback (RLHF), and multi-layered content filtering, we ensure generated outputs align with organizational values, regulatory requirements, and ethical guidelines.

Built for regulated industries requiring auditable AI systems, our platform provides comprehensive provenance tracking, watermarking, bias detection, and explainability. Every generation includes confidence scores, source attribution, and reasoning traces enabling human oversight and regulatory compliance. On-premise deployment options ensure sensitive data and proprietary workflows remain within your infrastructure.

Foundation Models

Custom GPT-style transformers for text generation and reasoning
Code generation models supporting 20+ programming languages
Diffusion models for high-fidelity image and video synthesis
Multimodal models processing text, images, audio, video
Efficient fine-tuning with LoRA, QLoRA, and parameter-efficient methods

Safety and Alignment

RLHF and constitutional AI for value alignment
Multi-layered content filtering for harmful content detection
Red team testing against adversarial attacks and jailbreaks
Bias detection and mitigation across demographics and domains
Hallucination detection with grounding and factuality verification

Enterprise Deployment

On-premise and air-gapped deployment options
GPU-optimized inference with TensorRT, vLLM, and quantization
Horizontal scaling handling millions of requests per day
A/B testing, canary deployments, and model versioning
Comprehensive monitoring, alerting, and SLA guarantees

Governance and Compliance

Model cards documenting capabilities, limitations, risks
Provenance tracking for training data and generated outputs
Watermarking and synthetic content detection
Audit logging for compliance and regulatory scrutiny
Human-in-the-loop workflows for high-stakes decisions

Customization and Fine-tuning

Domain-specific fine-tuning for legal, medical, financial text
Instruction tuning for task-specific performance
Few-shot and zero-shot learning capabilities
Retrieval-augmented generation (RAG) for grounded outputs
Custom knowledge base integration and semantic search

Developer Experience

OpenAI-compatible API for drop-in replacement
SDKs for Python, JavaScript, Java, Go
Streaming responses and batch processing endpoints
Fine-tuning API and model management dashboard
Comprehensive documentation, examples, and tutorials

Join the Generative AI Team

Build responsible foundation models and safety systems at the forefront of AI alignment and governance

Hiring planned → Coming soon

Principal Research Scientist, Foundation Models

Location: Philadelphia, PA or Remote

Compensation: $250,000 - $350,000 + equity

Type: Full-time

Lead research advancing state-of-the-art in large language models, diffusion models, and multimodal generation. Design novel architectures, scaling laws, and training methodologies while publishing at top-tier venues and deploying models serving enterprise customers.

Core Responsibilities

Design and train large-scale foundation models (100B+ parameters) for text, code, image, and multimodal generation using cutting-edge architectures
Develop novel pre-training objectives, scaling laws, and efficient training techniques reducing compute costs while improving model capabilities
Conduct original research publishable at NeurIPS, ICML, ICLR, ACL, CVPR advancing generative AI theory and practice
Build alignment techniques including RLHF, constitutional AI, and preference learning ensuring models follow human values and organizational policies
Optimize inference efficiency through quantization, pruning, distillation, and efficient attention mechanisms achieving sub-200ms latency
Collaborate with safety team to red team models, detect failure modes, and implement robust guardrails against adversarial attacks
Mentor PhD-level researchers and establish research culture balancing scientific rigor with rapid productionization

Required Qualifications

PhD in Computer Science, Machine Learning, or related field with 8+ years of deep learning research experience
Strong publication record at top-tier ML conferences (NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR) with 15+ peer-reviewed papers
Deep expertise in transformer architectures, attention mechanisms, and large-scale distributed training
Proven track record training foundation models at scale (10B+ parameters) using multi-GPU and multi-node clusters
Expert-level proficiency in PyTorch with experience in DeepSpeed, Megatron-LM, or similar distributed training frameworks
Strong understanding of scaling laws, emergent capabilities, and optimization techniques for large models
Experience deploying generative models to production serving millions of users

Preferred Qualifications

First-author publications specifically on LLMs, diffusion models, or multimodal foundation models
Experience with RLHF, instruction tuning, or constitutional AI for model alignment
Background in efficient transformers (Flash Attention, sparse attention, linear attention)
Contributions to open-source LLM projects (LLaMA, Mistral, Stable Diffusion, HuggingFace)
Postdoctoral research or experience at top AI labs (OpenAI, Anthropic, Google DeepMind, Meta FAIR)
Understanding of AI safety, interpretability, and responsible AI deployment

Why This Role Matters

Your research will advance the frontier of generative AI while ensuring safety, alignment, and responsible deployment. Publish groundbreaking work while building foundation models trusted by Fortune 500 enterprises and regulated industries.

Express Interest

Senior AI Safety Engineer

Location: Philadelphia, PA or Remote

Compensation: $185,000 - $260,000 + equity

Type: Full-time

Build safety systems and alignment infrastructure ensuring generative AI models are robust, controllable, and aligned with human values. Design red team testing frameworks, content filtering pipelines, and monitoring systems detecting harmful outputs and adversarial attacks.

Core Responsibilities

Design and implement multi-layered safety systems including content classifiers, toxicity detectors, and PII redaction pipelines
Build red team testing infrastructure probing models for vulnerabilities, jailbreaks, prompt injection attacks, and adversarial inputs
Develop hallucination detection systems using grounding, factuality verification, and retrieval-augmented generation techniques
Implement bias detection and mitigation strategies ensuring fairness across demographics, cultures, and sensitive attributes
Create monitoring dashboards tracking model behavior, safety metrics, and anomaly detection in production deployments
Collaborate with policy and legal teams to ensure compliance with AI regulations (EU AI Act, NIST AI RMF, industry standards)
Conduct safety evaluations, document risks, and communicate findings to leadership, customers, and regulators

Required Qualifications

BS/MS in Computer Science, Machine Learning, or related field with 6+ years of AI safety, adversarial ML, or trust and safety engineering
Deep expertise in content moderation, toxicity detection, and harmful content classification using ML techniques
Strong understanding of adversarial machine learning including prompt injection, jailbreaks, data poisoning, and model extraction
Experience building safety systems for production AI products serving millions of users
Proficiency in PyTorch, scikit-learn, and production ML frameworks (TensorFlow Serving, TorchServe)
Knowledge of bias detection methodologies, fairness metrics, and demographic parity evaluation
Familiarity with AI governance frameworks (NIST AI RMF, ISO 42001, EU AI Act requirements)

Preferred Qualifications

PhD in AI Safety, Machine Learning Security, or related research area
Publications on adversarial robustness, AI alignment, or safety evaluation benchmarks
Experience with constitutional AI, RLHF, or preference learning for value alignment
Background in watermarking, synthetic content detection, and provenance tracking
Contributions to AI safety organizations (Anthropic, OpenAI Safety, Center for AI Safety)
Understanding of interpretability techniques (SHAP, attention visualization, mechanistic interpretability)

Why This Role Matters

Your work ensures generative AI is deployed safely and responsibly, protecting users from harmful content, bias, and manipulation. Build safety infrastructure enabling enterprises to trust and adopt AI at scale while meeting regulatory requirements.

Express Interest

Director of Product, Generative AI

Location: Philadelphia, PA

Compensation: $200,000 - $280,000 + equity

Type: Full-time

Own product strategy and roadmap for Generative AI Platform, driving enterprise adoption of foundation models with built-in safety, governance, and compliance. Translate cutting-edge research into products generating measurable business value and competitive advantage for Fortune 500 customers.

Core Responsibilities

Define and execute product strategy for Generative AI Platform, balancing innovation with enterprise requirements for safety, governance, and auditability
Own P&L for Generative AI product line including revenue targets, pricing strategy, unit economics, and customer acquisition costs
Lead customer discovery with Fortune 500 enterprises across legal, healthcare, financial services, and regulated industries
Collaborate with research and safety teams to productize foundation models ensuring responsible deployment and regulatory compliance
Build product roadmap prioritizing model capabilities, API features, safety controls, enterprise integrations, and governance tools
Drive go-to-market execution partnering with sales, marketing, and customer success to achieve ARR growth and net revenue retention
Establish product metrics measuring model performance, safety accuracy, customer ROI, and competitive differentiation
Represent platform at AI conferences, engage with regulators, and influence industry standards for responsible AI deployment

Required Qualifications

8+ years of product management experience with 5+ years in B2B enterprise AI/ML platforms or generative AI products
Proven track record owning P&L and driving $20M+ ARR growth in technical AI products serving Fortune 500 enterprises
Deep understanding of generative AI, LLMs, diffusion models, and enterprise deployment challenges
Technical fluency to engage with AI researchers and engineers, understanding model architectures and training methodologies
Experience navigating AI governance, responsible AI frameworks, and regulatory requirements (EU AI Act, sector-specific regulations)
Strong quantitative skills with expertise in product analytics, customer ROI analysis, and competitive positioning
Executive presence to engage C-level buyers, present at industry conferences, and represent company to regulators and policymakers

Preferred Qualifications

MBA from top-tier program or equivalent strategic business experience
Background in AI platforms (OpenAI, Anthropic, Cohere, Google Vertex AI) or enterprise ML infrastructure
Experience with government or highly regulated industry customers requiring FedRAMP, HIPAA, or similar certifications
Technical degree in Computer Science, AI, or quantitative field with hands-on ML experience
Track record building 0-to-1 AI products addressing emerging needs in content generation, code assistance, or creative workflows
Network in AI community through conferences (NeurIPS, ICML), industry groups, or AI safety organizations

Why This Role Matters

Shape the future of enterprise generative AI, enabling responsible deployment of foundation models at Fortune 500 scale. Build products balancing innovation with safety, transforming how organizations leverage AI for content creation, decision support, and knowledge work.

Express Interest