Tech-tag co-occurrence

Every AI role gets tagged from a 35-term controlled vocabulary spanning agent / eval / training / inference / modality clusters. Tags that frequently appear together in the same JD pull each other close; thicker edges = more co-occurrences. The clusters that emerge organically are the real sub-disciplines of applied AI engineering right now.

All sectors · 4742Enterprise · 1275AI Frontier · 868Industrial · 650Data AI · 526Consumer · 337Banking · 239Fintech · 220Robotics · 127Defense · 115Pharma · 101Media · 73Retail · 62Hospitality · 62Telecom · 27Seattle · 27Aerospace · 14Insurance · 12Healthtech · 7

Showing 526 tagged AI roles in Data AI. Layout is a Fruchterman-Reingold force simulation, run server-side to convergence.

fine_tuning ↔ frontier_research (11 JDs)frontier_research ↔ inference_infra (7 JDs)frontier_research ↔ model_serving (10 JDs)fine_tuning ↔ inference_infra (24 JDs)fine_tuning ↔ model_serving (62 JDs)inference_infra ↔ model_serving (152 JDs)agent_orchestration ↔ inference_infra (49 JDs)agent_orchestration ↔ model_serving (116 JDs)fine_tuning ↔ pretraining (3 JDs)inference_infra ↔ pretraining (5 JDs)model_serving ↔ pretraining (6 JDs)agent_orchestration ↔ agent_research (29 JDs)agent_orchestration ↔ evals (82 JDs)agent_orchestration ↔ llm_observability (166 JDs)agent_orchestration ↔ guardrails (76 JDs)evals ↔ llm_observability (70 JDs)evals ↔ guardrails (26 JDs)evals ↔ model_serving (60 JDs)guardrails ↔ llm_observability (75 JDs)llm_observability ↔ model_serving (105 JDs)guardrails ↔ model_serving (32 JDs)evals ↔ rl_post_training (18 JDs)agent_research ↔ evals (20 JDs)evals ↔ frontier_research (15 JDs)agent_research ↔ rl_post_training (13 JDs)frontier_research ↔ rl_post_training (9 JDs)model_serving ↔ rl_post_training (11 JDs)agent_research ↔ frontier_research (5 JDs)agent_research ↔ model_serving (9 JDs)agent_orchestration ↔ fine_tuning (57 JDs)agent_research ↔ fine_tuning (10 JDs)agent_research ↔ inference_infra (4 JDs)evals ↔ fine_tuning (39 JDs)evals ↔ inference_infra (25 JDs)agent_orchestration ↔ multimodal (20 JDs)model_serving ↔ multimodal (11 JDs)inference_infra ↔ multimodal (6 JDs)fine_tuning ↔ multimodal (13 JDs)agent_orchestration ↔ tool_use (84 JDs)agent_orchestration ↔ recommender_systems (8 JDs)model_serving ↔ tool_use (31 JDs)inference_infra ↔ tool_use (14 JDs)recommender_systems ↔ tool_use (3 JDs)multimodal ↔ tool_use (11 JDs)model_serving ↔ recommender_systems (10 JDs)inference_infra ↔ recommender_systems (6 JDs)multimodal ↔ recommender_systems (2 JDs)fine_tuning ↔ tool_use (14 JDs)agent_orchestration ↔ rag (131 JDs)agent_orchestration ↔ vector_db (46 JDs)llm_observability ↔ rag (92 JDs)llm_observability ↔ vector_db (35 JDs)rag ↔ vector_db (54 JDs)model_serving ↔ rag (64 JDs)model_serving ↔ vector_db (42 JDs)inference_infra ↔ llm_observability (50 JDs)recommender_systems ↔ search_ranking (11 JDs)llm_observability ↔ tool_use (62 JDs)guardrails ↔ tool_use (36 JDs)agent_research ↔ llm_observability (18 JDs)fine_tuning ↔ llm_observability (42 JDs)inference_infra ↔ vision (2 JDs)audio_speech ↔ inference_infra (3 JDs)model_serving ↔ training_infra (2 JDs)model_serving ↔ vision (4 JDs)audio_speech ↔ model_serving (3 JDs)llm_observability ↔ vision (2 JDs)rag ↔ tool_use (32 JDs)tool_use ↔ vector_db (7 JDs)evals ↔ tool_use (16 JDs)guardrails ↔ rag (39 JDs)guardrails ↔ vector_db (11 JDs)guardrails ↔ inference_infra (11 JDs)evals ↔ rag (50 JDs)inference_infra ↔ rag (24 JDs)evals ↔ vector_db (21 JDs)inference_infra ↔ vector_db (28 JDs)llm_observability ↔ recommender_systems (9 JDs)fine_tuning ↔ rag (45 JDs)rag ↔ synthetic_data (3 JDs)fine_tuning ↔ synthetic_data (7 JDs)agent_orchestration ↔ synthetic_data (6 JDs)evals ↔ synthetic_data (13 JDs)agent_orchestration ↔ vision (3 JDs)fine_tuning ↔ vector_db (14 JDs)forecasting ↔ recommender_systems (2 JDs)multimodal ↔ rag (9 JDs)fine_tuning ↔ guardrails (15 JDs)llm_observability ↔ multimodal (11 JDs)guardrails ↔ multimodal (7 JDs)evals ↔ multimodal (8 JDs)agent_orchestration ↔ frontier_research (3 JDs)agent_orchestration ↔ rl_post_training (9 JDs)agent_research ↔ guardrails (10 JDs)frontier_research ↔ guardrails (2 JDs)guardrails ↔ rl_post_training (5 JDs)frontier_research ↔ synthetic_data (4 JDs)frontier_research ↔ llm_observability (4 JDs)agent_research ↔ multimodal (5 JDs)fine_tuning ↔ vision (2 JDs)multimodal ↔ vision (6 JDs)agent_orchestration ↔ code_gen (4 JDs)code_gen ↔ tool_use (3 JDs)code_gen ↔ llm_observability (3 JDs)code_gen ↔ rag (3 JDs)agent_orchestration ↔ search_ranking (4 JDs)evals ↔ search_ranking (9 JDs)llm_observability ↔ search_ranking (9 JDs)model_serving ↔ search_ranking (7 JDs)inference_infra ↔ search_ranking (3 JDs)agent_research ↔ tool_use (7 JDs)evals ↔ recommender_systems (7 JDs)agent_research ↔ rag (4 JDs)rag ↔ search_ranking (11 JDs)search_ranking ↔ vector_db (6 JDs)rag ↔ recommender_systems (8 JDs)recommender_systems ↔ vector_db (6 JDs)rl_post_training ↔ synthetic_data (5 JDs)agent_research ↔ synthetic_data (2 JDs)multimodal ↔ synthetic_data (2 JDs)multimodal ↔ rl_post_training (4 JDs)frontier_research ↔ multimodal (4 JDs)embodied_ai ↔ model_serving (2 JDs)embodied_ai ↔ fine_tuning (2 JDs)inference_infra ↔ quantization (2 JDs)model_serving ↔ quantization (2 JDs)frontier_research ↔ pretraining (4 JDs)model_serving ↔ synthetic_data (2 JDs)fine_tuning ↔ search_ranking (2 JDs)fine_tuning ↔ recommender_systems (2 JDs)reward_modeling ↔ rl_post_training (2 JDs)fine_tuning ↔ rl_post_training (15 JDs)fine_tuning ↔ reward_modeling (2 JDs)agent_research ↔ vector_db (2 JDs)inference_infra ↔ rl_post_training (9 JDs)llm_observability ↔ synthetic_data (2 JDs)llm_observability ↔ rl_post_training (6 JDs)rl_robotics ↔ synthetic_data (2 JDs)evals ↔ rl_robotics (2 JDs)multimodal ↔ rl_robotics (2 JDs)rl_post_training ↔ tool_use (2 JDs)embodied_ai ↔ rl_robotics (2 JDs)guardrails ↔ search_ranking (2 JDs)Frontier research N=24 JDs Top co-occur: Evals ×15 · Fine-tuning ×11 · Model serving ×10Fine-tuning N=98 JDs Top co-occur: Model serving ×62 · Agent orchestration ×57 · RAG ×45Fine-tuningInference infra N=153 JDs Top co-occur: Model serving ×152 · LLM observability ×50 · Agent orchestration ×49Inference infraModel serving N=254 JDs Top co-occur: Inference infra ×152 · Agent orchestration ×116 · LLM observability ×105Model servingAgent orchestration N=317 JDs Top co-occur: LLM observability ×166 · RAG ×131 · Model serving ×116Agent orchestrationPretraining N=7 JDs Top co-occur: Model serving ×6 · Inference infra ×5 · Frontier research ×4Agent research N=37 JDs Top co-occur: Agent orchestration ×29 · Evals ×20 · LLM observability ×18Evals N=126 JDs Top co-occur: Agent orchestration ×82 · LLM observability ×70 · Model serving ×60EvalsLLM observability N=223 JDs Top co-occur: Agent orchestration ×166 · Model serving ×105 · RAG ×92LLM observabilityGuardrails N=81 JDs Top co-occur: Agent orchestration ×76 · LLM observability ×75 · RAG ×39GuardrailsRL post-training N=30 JDs Top co-occur: Evals ×18 · Fine-tuning ×15 · Agent research ×13Multimodal N=30 JDs Top co-occur: Agent orchestration ×20 · Fine-tuning ×13 · Model serving ×11Tool use N=85 JDs Top co-occur: Agent orchestration ×84 · LLM observability ×62 · Guardrails ×36Tool useRecommender systems N=21 JDs Top co-occur: Search & ranking ×11 · Model serving ×10 · LLM observability ×9RAG N=158 JDs Top co-occur: Agent orchestration ×131 · LLM observability ×92 · Model serving ×64RAGVector DB N=64 JDs Top co-occur: RAG ×54 · Agent orchestration ×46 · Model serving ×42Search & ranking N=15 JDs Top co-occur: Recommender systems ×11 · RAG ×11 · Evals ×9Training infra N=4 JDs Top co-occur: Model serving ×2Vision N=9 JDs Top co-occur: Multimodal ×6 · Model serving ×4 · Agent orchestration ×3Audio & speech N=3 JDs Top co-occur: Inference infra ×3 · Model serving ×3Synthetic data N=17 JDs Top co-occur: Evals ×13 · Fine-tuning ×7 · Agent orchestration ×6Forecasting N=4 JDs Top co-occur: Recommender systems ×2Code gen N=5 JDs Top co-occur: Agent orchestration ×4 · Tool use ×3 · LLM observability ×3Data pipeline N=1 JDsEmbodied AI N=5 JDs Top co-occur: Model serving ×2 · Fine-tuning ×2 · RL robotics ×2Quantization N=2 JDs Top co-occur: Inference infra ×2 · Model serving ×2Multi-agent N=1 JDsSemantic search N=1 JDsReward modeling N=2 JDs Top co-occur: RL post-training ×2 · Fine-tuning ×2RLHF N=3 JDsRL robotics N=3 JDs Top co-occur: Synthetic data ×2 · Evals ×2 · Multimodal ×2Interpretability N=1 JDs
32 tags · 143 co-occurrence edges · min edge weight 2. Bubble area ∝ JDs containing tag · edge thickness ∝ co-occurrence count. Hover any node for top-3 partners; click to see the JDs.

Tag velocity · last 4 weeks vs prior 4

Which technologies are hot, which are cooling. Sparkline = 12 weeks of unique roles tagged with each term, last bar on the right is this week. Sorted by absolute pickup. Tags with under 10 lifetime mentions are hidden as noise.

Tag12-week trendLast 4wPrior 4wΔTotal
agent_orchestration
7 roles
Feb 23
14 roles
Mar 2
18 roles
9
9 roles
16
13 roles
23
22 roles
30
12 roles
Apr 6
36 roles
13
29 roles
20
22 roles
27
28 roles
May 4
0 roles
11
11556↑+59317
model_serving
8 roles
Feb 23
4 roles
Mar 2
10 roles
9
2 roles
16
10 roles
23
14 roles
30
5 roles
Apr 6
23 roles
13
14 roles
20
18 roles
27
19 roles
May 4
0 roles
11
7431↑+43254
llm_observability
6 roles
Feb 23
12 roles
Mar 2
14 roles
9
2 roles
16
12 roles
23
15 roles
30
4 roles
Apr 6
26 roles
13
22 roles
20
12 roles
27
16 roles
May 4
0 roles
11
7633↑+43223
evals
5 roles
Feb 23
3 roles
Mar 2
1 role
9
4 roles
16
8 roles
23
1 role
30
1 role
Apr 6
17 roles
13
9 roles
20
8 roles
27
7 roles
May 4
0 roles
11
4114↑+27126
tool_use
1 role
Feb 23
5 roles
Mar 2
9 roles
9
1 role
16
4 roles
23
2 roles
30
3 roles
Apr 6
9 roles
13
13 roles
20
3 roles
27
10 roles
May 4
0 roles
11
3510↑+2585
inference_infra
6 roles
Feb 23
2 roles
Mar 2
4 roles
9
2 roles
16
8 roles
23
9 roles
30
4 roles
Apr 6
11 roles
13
5 roles
20
11 roles
27
12 roles
May 4
0 roles
11
3923↑+16153
fine_tuning
3 roles
Feb 23
4 roles
Mar 2
2 roles
9
2 roles
16
6 roles
23
2 roles
30
1 role
Apr 6
6 roles
13
10 roles
20
2 roles
27
9 roles
May 4
0 roles
11
2711↑+1698
guardrails
4 roles
Feb 23
4 roles
Mar 2
11 roles
9
2 roles
16
6 roles
23
7 roles
30
1 role
Apr 6
12 roles
13
9 roles
20
3 roles
27
5 roles
May 4
0 roles
11
2916↑+1381
rag
6 roles
Feb 23
5 roles
Mar 2
10 roles
9
2 roles
16
9 roles
23
14 roles
30
9 roles
Apr 6
13 roles
13
12 roles
20
10 roles
27
10 roles
May 4
0 roles
11
4435↑+9158
multimodal
1 role
Feb 23
0 roles
Mar 2
0 roles
9
0 roles
16
3 roles
23
0 roles
30
1 role
Apr 6
0 roles
13
2 roles
20
1 role
27
6 roles
May 4
0 roles
11
94↑+530
vector_db
3 roles
Feb 23
1 role
Mar 2
1 role
9
1 role
16
5 roles
23
7 roles
30
1 role
Apr 6
9 roles
13
3 roles
20
3 roles
27
2 roles
May 4
0 roles
11
1714↑+364
agent_research
1 role
Feb 23
2 roles
Mar 2
2 roles
9
2 roles
16
3 roles
23
1 role
30
1 role
Apr 6
4 roles
13
3 roles
20
1 role
27
2 roles
May 4
0 roles
11
107↑+337
frontier_research
1 role
Feb 23
0 roles
Mar 2
0 roles
9
0 roles
16
2 roles
23
1 role
30
0 roles
Apr 6
1 role
13
0 roles
20
2 roles
27
3 roles
May 4
0 roles
11
63↑+324
synthetic_data
1 role
Feb 23
1 role
Mar 2
0 roles
9
1 role
16
1 role
23
0 roles
30
0 roles
Apr 6
4 roles
13
0 roles
20
1 role
27
0 roles
May 4
0 roles
11
52↑+317
recommender_systems
0 roles
Feb 23
1 role
Mar 2
0 roles
9
0 roles
16
0 roles
23
1 role
30
1 role
Apr 6
1 role
13
1 role
20
0 roles
27
2 roles
May 4
0 roles
11
42↑+221
rl_post_training
2 roles
Feb 23
1 role
Mar 2
0 roles
9
2 roles
16
2 roles
23
1 role
30
0 roles
Apr 6
2 roles
13
1 role
20
0 roles
27
3 roles
May 4
0 roles
11
65↑+130
search_ranking
2 roles
Feb 23
0 roles
Mar 2
0 roles
9
0 roles
16
1 role
23
1 role
30
0 roles
Apr 6
0 roles
13
0 roles
20
0 roles
27
2 roles
May 4
0 roles
11
22·015