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 1275 tagged AI roles in Enterprise. Layout is a Fruchterman-Reingold force simulation, run server-side to convergence.

agent_orchestration ↔ rag (310 JDs)agent_orchestration ↔ vector_db (101 JDs)agent_orchestration ↔ fine_tuning (146 JDs)agent_orchestration ↔ model_serving (339 JDs)agent_orchestration ↔ inference_infra (120 JDs)agent_orchestration ↔ llm_observability (433 JDs)agent_orchestration ↔ evals (184 JDs)rag ↔ vector_db (122 JDs)fine_tuning ↔ rag (116 JDs)model_serving ↔ rag (190 JDs)inference_infra ↔ rag (67 JDs)llm_observability ↔ rag (223 JDs)evals ↔ rag (82 JDs)fine_tuning ↔ vector_db (41 JDs)model_serving ↔ vector_db (73 JDs)inference_infra ↔ vector_db (36 JDs)llm_observability ↔ vector_db (72 JDs)evals ↔ vector_db (23 JDs)fine_tuning ↔ model_serving (189 JDs)fine_tuning ↔ inference_infra (55 JDs)fine_tuning ↔ llm_observability (110 JDs)evals ↔ fine_tuning (73 JDs)inference_infra ↔ model_serving (264 JDs)llm_observability ↔ model_serving (249 JDs)evals ↔ model_serving (121 JDs)inference_infra ↔ llm_observability (88 JDs)evals ↔ inference_infra (42 JDs)evals ↔ llm_observability (136 JDs)agent_orchestration ↔ tool_use (244 JDs)agent_orchestration ↔ guardrails (165 JDs)rag ↔ tool_use (94 JDs)guardrails ↔ rag (64 JDs)tool_use ↔ vector_db (22 JDs)guardrails ↔ vector_db (19 JDs)fine_tuning ↔ tool_use (36 JDs)fine_tuning ↔ guardrails (44 JDs)model_serving ↔ tool_use (82 JDs)guardrails ↔ model_serving (73 JDs)inference_infra ↔ tool_use (28 JDs)guardrails ↔ inference_infra (33 JDs)llm_observability ↔ tool_use (134 JDs)guardrails ↔ llm_observability (129 JDs)evals ↔ tool_use (69 JDs)evals ↔ guardrails (90 JDs)guardrails ↔ tool_use (67 JDs)agent_orchestration ↔ agent_research (94 JDs)agent_research ↔ tool_use (21 JDs)agent_research ↔ evals (35 JDs)agent_research ↔ guardrails (30 JDs)agent_research ↔ fine_tuning (33 JDs)agent_research ↔ inference_infra (9 JDs)agent_research ↔ model_serving (39 JDs)agent_research ↔ rag (25 JDs)multimodal ↔ vision (37 JDs)fine_tuning ↔ vision (30 JDs)fine_tuning ↔ multimodal (74 JDs)model_serving ↔ recommender_systems (43 JDs)model_serving ↔ multi_agent (3 JDs)fine_tuning ↔ recommender_systems (24 JDs)fine_tuning ↔ multi_agent (2 JDs)agent_orchestration ↔ multimodal (53 JDs)multimodal ↔ rag (26 JDs)multimodal ↔ vector_db (11 JDs)model_serving ↔ multimodal (76 JDs)agent_orchestration ↔ recommender_systems (37 JDs)agent_research ↔ recommender_systems (7 JDs)agent_research ↔ multimodal (20 JDs)rag ↔ recommender_systems (28 JDs)multimodal ↔ recommender_systems (8 JDs)agent_research ↔ llm_observability (61 JDs)fine_tuning ↔ pretraining (13 JDs)multimodal ↔ pretraining (15 JDs)pretraining ↔ vision (8 JDs)inference_infra ↔ multimodal (25 JDs)llm_observability ↔ multimodal (42 JDs)evals ↔ multimodal (32 JDs)rag ↔ semantic_search (4 JDs)semantic_search ↔ vector_db (3 JDs)llm_observability ↔ semantic_search (4 JDs)agent_orchestration ↔ code_gen (40 JDs)code_gen ↔ tool_use (18 JDs)code_gen ↔ llm_observability (24 JDs)rag ↔ search_ranking (31 JDs)search_ranking ↔ vector_db (23 JDs)llm_observability ↔ search_ranking (16 JDs)audio_speech ↔ llm_observability (4 JDs)audio_speech ↔ model_serving (11 JDs)agent_orchestration ↔ frontier_research (10 JDs)evals ↔ frontier_research (10 JDs)agent_orchestration ↔ synthetic_data (8 JDs)synthetic_data ↔ tool_use (2 JDs)model_serving ↔ synthetic_data (5 JDs)evals ↔ synthetic_data (8 JDs)rag ↔ synthetic_data (2 JDs)agent_orchestration ↔ search_ranking (23 JDs)fine_tuning ↔ search_ranking (12 JDs)model_serving ↔ search_ranking (32 JDs)recommender_systems ↔ search_ranking (28 JDs)llm_observability ↔ recommender_systems (24 JDs)inference_infra ↔ training_infra (2 JDs)model_serving ↔ training_infra (2 JDs)frontier_research ↔ llm_observability (5 JDs)fine_tuning ↔ frontier_research (19 JDs)frontier_research ↔ model_serving (15 JDs)agent_research ↔ frontier_research (9 JDs)code_gen ↔ evals (13 JDs)agent_research ↔ code_gen (4 JDs)recommender_systems ↔ vector_db (13 JDs)agent_research ↔ vector_db (8 JDs)evals ↔ recommender_systems (7 JDs)guardrails ↔ recommender_systems (4 JDs)recommender_systems ↔ tool_use (4 JDs)multimodal ↔ search_ranking (5 JDs)evals ↔ search_ranking (5 JDs)guardrails ↔ search_ranking (3 JDs)search_ranking ↔ tool_use (4 JDs)agent_research ↔ search_ranking (5 JDs)guardrails ↔ multimodal (10 JDs)multimodal ↔ tool_use (16 JDs)fine_tuning ↔ synthetic_data (10 JDs)code_gen ↔ guardrails (4 JDs)code_gen ↔ model_serving (13 JDs)frontier_research ↔ multimodal (21 JDs)audio_speech ↔ multimodal (14 JDs)model_serving ↔ pretraining (6 JDs)frontier_research ↔ pretraining (8 JDs)audio_speech ↔ pretraining (5 JDs)audio_speech ↔ fine_tuning (12 JDs)model_serving ↔ vision (24 JDs)frontier_research ↔ vision (6 JDs)audio_speech ↔ frontier_research (7 JDs)audio_speech ↔ vision (13 JDs)code_gen ↔ fine_tuning (6 JDs)model_serving ↔ quantization (3 JDs)inference_infra ↔ quantization (3 JDs)fine_tuning ↔ quantization (2 JDs)multimodal ↔ quantization (3 JDs)multi_agent ↔ rag (3 JDs)agent_orchestration ↔ multi_agent (10 JDs)evals ↔ multi_agent (3 JDs)agent_research ↔ vision (3 JDs)evals ↔ vision (9 JDs)synthetic_data ↔ vision (5 JDs)multimodal ↔ synthetic_data (12 JDs)agent_research ↔ synthetic_data (3 JDs)agent_orchestration ↔ audio_speech (4 JDs)audio_speech ↔ evals (2 JDs)agent_orchestration ↔ vision (7 JDs)tool_use ↔ vision (2 JDs)llm_observability ↔ vision (8 JDs)fine_tuning ↔ rl_post_training (18 JDs)frontier_research ↔ rl_post_training (9 JDs)multimodal ↔ rl_post_training (12 JDs)multi_agent ↔ tool_use (2 JDs)inference_infra ↔ recommender_systems (19 JDs)llm_observability ↔ multi_agent (3 JDs)code_gen ↔ rag (5 JDs)code_gen ↔ vector_db (2 JDs)pretraining ↔ synthetic_data (2 JDs)inference_infra ↔ search_ranking (18 JDs)code_gen ↔ inference_infra (4 JDs)agent_orchestration ↔ embodied_ai (2 JDs)embodied_ai ↔ model_serving (2 JDs)embodied_ai ↔ inference_infra (2 JDs)audio_speech ↔ synthetic_data (3 JDs)evals ↔ rl_post_training (10 JDs)frontier_research ↔ synthetic_data (3 JDs)inference_infra ↔ vision (5 JDs)forecasting ↔ rag (4 JDs)forecasting ↔ llm_observability (2 JDs)fine_tuning ↔ forecasting (3 JDs)rag ↔ vision (2 JDs)vector_db ↔ vision (2 JDs)recommender_systems ↔ vision (2 JDs)search_ranking ↔ vision (2 JDs)multimodal ↔ rlhf (7 JDs)frontier_research ↔ rlhf (3 JDs)frontier_research ↔ inference_infra (5 JDs)rl_post_training ↔ rlhf (3 JDs)agent_orchestration ↔ rl_post_training (13 JDs)model_serving ↔ rl_post_training (15 JDs)inference_infra ↔ rl_post_training (4 JDs)rl_post_training ↔ synthetic_data (3 JDs)agent_orchestration ↔ rlhf (7 JDs)model_serving ↔ rlhf (6 JDs)inference_infra ↔ rlhf (2 JDs)evals ↔ rlhf (6 JDs)rlhf ↔ synthetic_data (3 JDs)inference_infra ↔ synthetic_data (2 JDs)audio_speech ↔ inference_infra (3 JDs)agent_orchestration ↔ semantic_search (3 JDs)agent_orchestration ↔ forecasting (2 JDs)agent_research ↔ rl_post_training (8 JDs)rl_post_training ↔ tool_use (3 JDs)llm_observability ↔ rl_post_training (7 JDs)rag ↔ rl_post_training (2 JDs)recommender_systems ↔ rl_post_training (3 JDs)frontier_research ↔ rag (3 JDs)guardrails ↔ rl_post_training (3 JDs)fine_tuning ↔ interpretability (2 JDs)interpretability ↔ model_serving (2 JDs)frontier_research ↔ guardrails (2 JDs)frontier_research ↔ interpretability (2 JDs)guardrails ↔ interpretability (2 JDs)rl_post_training ↔ search_ranking (2 JDs)agent_research ↔ rlhf (4 JDs)fine_tuning ↔ rlhf (5 JDs)llm_observability ↔ rlhf (5 JDs)agent_orchestration ↔ reward_modeling (3 JDs)fine_tuning ↔ reward_modeling (3 JDs)model_serving ↔ reward_modeling (4 JDs)multimodal ↔ reward_modeling (4 JDs)evals ↔ reward_modeling (3 JDs)llm_observability ↔ synthetic_data (3 JDs)frontier_research ↔ reward_modeling (3 JDs)agent_research ↔ reward_modeling (2 JDs)reward_modeling ↔ rl_post_training (4 JDs)reward_modeling ↔ rlhf (2 JDs)rlhf ↔ tool_use (2 JDs)Agent orchestration N=899 JDs Top co-occur: LLM observability ×433 · Model serving ×339 · RAG ×310Agent orchestrationRAG N=364 JDs Top co-occur: Agent orchestration ×310 · LLM observability ×223 · Model serving ×190RAGVector DB N=129 JDs Top co-occur: RAG ×122 · Agent orchestration ×101 · Model serving ×73Fine-tuning N=255 JDs Top co-occur: Model serving ×189 · Agent orchestration ×146 · RAG ×116Fine-tuningModel serving N=584 JDs Top co-occur: Agent orchestration ×339 · Inference infra ×264 · LLM observability ×249Model servingInference infra N=267 JDs Top co-occur: Model serving ×264 · Agent orchestration ×120 · LLM observability ×88Inference infraLLM observability N=514 JDs Top co-occur: Agent orchestration ×433 · Model serving ×249 · RAG ×223LLM observabilityEvals N=229 JDs Top co-occur: Agent orchestration ×184 · LLM observability ×136 · Model serving ×121EvalsTool use N=244 JDs Top co-occur: Agent orchestration ×244 · LLM observability ×134 · RAG ×94Tool useGuardrails N=193 JDs Top co-occur: Agent orchestration ×165 · LLM observability ×129 · Evals ×90Agent research N=99 JDs Top co-occur: Agent orchestration ×94 · LLM observability ×61 · Model serving ×39Vision N=41 JDs Top co-occur: Multimodal ×37 · Fine-tuning ×30 · Model serving ×24Multimodal N=141 JDs Top co-occur: Model serving ×76 · Fine-tuning ×74 · Agent orchestration ×53Recommender systems N=71 JDs Top co-occur: Model serving ×43 · Agent orchestration ×37 · RAG ×28Multi-agent N=11 JDs Top co-occur: Agent orchestration ×10 · Model serving ×3 · RAG ×3Forecasting N=7 JDs Top co-occur: RAG ×4 · Fine-tuning ×3 · LLM observability ×2Pretraining N=16 JDs Top co-occur: Multimodal ×15 · Fine-tuning ×13 · Vision ×8Semantic search N=6 JDs Top co-occur: RAG ×4 · LLM observability ×4 · Vector DB ×3Code gen N=45 JDs Top co-occur: Agent orchestration ×40 · LLM observability ×24 · Tool use ×18Search & ranking N=46 JDs Top co-occur: Model serving ×32 · RAG ×31 · Recommender systems ×28Audio & speech N=22 JDs Top co-occur: Multimodal ×14 · Vision ×13 · Fine-tuning ×12Frontier research N=29 JDs Top co-occur: Multimodal ×21 · Fine-tuning ×19 · Model serving ×15Synthetic data N=17 JDs Top co-occur: Multimodal ×12 · Fine-tuning ×10 · Agent orchestration ×8Training infra N=4 JDs Top co-occur: Inference infra ×2 · Model serving ×2Quantization N=3 JDs Top co-occur: Model serving ×3 · Inference infra ×3 · Multimodal ×3RL post-training N=25 JDs Top co-occur: Fine-tuning ×18 · Model serving ×15 · Agent orchestration ×13Embodied AI N=2 JDs Top co-occur: Agent orchestration ×2 · Model serving ×2 · Inference infra ×2RLHF N=7 JDs Top co-occur: Multimodal ×7 · Agent orchestration ×7 · Model serving ×6Interpretability N=2 JDs Top co-occur: Fine-tuning ×2 · Model serving ×2 · Frontier research ×2Reward modeling N=4 JDs Top co-occur: Model serving ×4 · Multimodal ×4 · RL post-training ×4RL robotics N=1 JDsData pipeline N=1 JDsDistillation N=1 JDs
33 tags · 219 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
13 roles
Feb 23
20 roles
Mar 2
42 roles
9
60 roles
16
42 roles
23
92 roles
30
35 roles
Apr 6
57 roles
13
85 roles
20
96 roles
27
125 roles
May 4
0 roles
11
353214↑+139899
llm_observability
6 roles
Feb 23
11 roles
Mar 2
19 roles
9
41 roles
16
28 roles
23
58 roles
30
23 roles
Apr 6
31 roles
13
44 roles
20
56 roles
27
69 roles
May 4
0 roles
11
194141↑+53514
model_serving
8 roles
Feb 23
11 roles
Mar 2
17 roles
9
36 roles
16
23 roles
23
81 roles
30
24 roles
Apr 6
32 roles
13
39 roles
20
62 roles
27
81 roles
May 4
0 roles
11
209157↑+52584
evals
4 roles
Feb 23
4 roles
Mar 2
6 roles
9
17 roles
16
10 roles
23
21 roles
30
8 roles
Apr 6
16 roles
13
18 roles
20
35 roles
27
31 roles
May 4
0 roles
11
9951↑+48229
tool_use
3 roles
Feb 23
2 roles
Mar 2
9 roles
9
20 roles
16
9 roles
23
22 roles
30
12 roles
Apr 6
17 roles
13
17 roles
20
41 roles
27
29 roles
May 4
0 roles
11
10160↑+41244
rag
6 roles
Feb 23
5 roles
Mar 2
15 roles
9
34 roles
16
18 roles
23
42 roles
30
12 roles
Apr 6
24 roles
13
22 roles
20
38 roles
27
53 roles
May 4
0 roles
11
13097↑+33364
guardrails
5 roles
Feb 23
7 roles
Mar 2
10 roles
9
13 roles
16
11 roles
23
14 roles
30
10 roles
Apr 6
11 roles
13
14 roles
20
22 roles
27
32 roles
May 4
0 roles
11
7846↑+32193
inference_infra
3 roles
Feb 23
4 roles
Mar 2
11 roles
9
12 roles
16
12 roles
23
46 roles
30
10 roles
Apr 6
18 roles
13
17 roles
20
24 roles
27
42 roles
May 4
0 roles
11
10077↑+23267
agent_research
3 roles
Feb 23
3 roles
Mar 2
8 roles
9
8 roles
16
3 roles
23
9 roles
30
1 role
Apr 6
3 roles
13
8 roles
20
8 roles
27
15 roles
May 4
0 roles
11
3319↑+1499
search_ranking
1 role
Feb 23
1 role
Mar 2
3 roles
9
0 roles
16
2 roles
23
1 role
30
0 roles
Apr 6
4 roles
13
4 roles
20
4 roles
27
4 roles
May 4
0 roles
11
163↑+1346
fine_tuning
2 roles
Feb 23
5 roles
Mar 2
8 roles
9
16 roles
16
10 roles
23
45 roles
30
11 roles
Apr 6
18 roles
13
23 roles
20
26 roles
27
27 roles
May 4
0 roles
11
9180↑+11255
vector_db
4 roles
Feb 23
2 roles
Mar 2
8 roles
9
9 roles
16
11 roles
23
14 roles
30
4 roles
Apr 6
4 roles
13
7 roles
20
14 roles
27
17 roles
May 4
0 roles
11
4232↑+10129
code_gen
1 role
Feb 23
2 roles
Mar 2
1 role
9
4 roles
16
2 roles
23
1 role
30
2 roles
Apr 6
1 role
13
3 roles
20
7 roles
27
7 roles
May 4
0 roles
11
189↑+945
recommender_systems
1 role
Feb 23
2 roles
Mar 2
1 role
9
2 roles
16
5 roles
23
9 roles
30
2 roles
Apr 6
5 roles
13
5 roles
20
3 roles
27
10 roles
May 4
0 roles
11
2317↑+671
multi_agent
0 roles
Feb 23
2 roles
Mar 2
0 roles
9
0 roles
16
1 role
23
1 role
30
0 roles
Apr 6
1 role
13
0 roles
20
2 roles
27
3 roles
May 4
0 roles
11
62↑+411
rl_post_training
3 roles
Feb 23
1 role
Mar 2
1 role
9
0 roles
16
1 role
23
2 roles
30
0 roles
Apr 6
1 role
13
0 roles
20
1 role
27
1 role
May 4
0 roles
11
33·025
synthetic_data
0 roles
Feb 23
0 roles
Mar 2
0 roles
9
1 role
16
1 role
23
5 roles
30
1 role
Apr 6
2 roles
13
1 role
20
3 roles
27
1 role
May 4
0 roles
11
77·017
audio_speech
0 roles
Feb 23
0 roles
Mar 2
1 role
9
0 roles
16
0 roles
23
10 roles
30
0 roles
Apr 6
1 role
13
3 roles
20
2 roles
27
3 roles
May 4
0 roles
11
910↓-122
frontier_research
3 roles
Feb 23
0 roles
Mar 2
1 role
9
2 roles
16
2 roles
23
7 roles
30
1 role
Apr 6
3 roles
13
0 roles
20
2 roles
27
0 roles
May 4
0 roles
11
511↓-629
pretraining
0 roles
Feb 23
0 roles
Mar 2
0 roles
9
3 roles
16
1 role
23
8 roles
30
0 roles
Apr 6
1 role
13
0 roles
20
1 role
27
1 role
May 4
0 roles
11
39↓-616
vision
0 roles
Feb 23
1 role
Mar 2
1 role
9
1 role
16
0 roles
23
19 roles
30
3 roles
Apr 6
2 roles
13
2 roles
20
4 roles
27
4 roles
May 4
0 roles
11
1222↓-1041
multimodal
3 roles
Feb 23
3 roles
Mar 2
2 roles
9
20 roles
16
6 roles
23
38 roles
30
5 roles
Apr 6
7 roles
13
10 roles
20
10 roles
27
16 roles
May 4
0 roles
11
4356↓-13141